LIGHT AND WORK OTHER BOOKS BY M. LUCKIESH 
Light and Color in Advertising and Merchandising 
Ultraviolet Radiation 
Visual Illusions; Their Causes, Characteristics and 
Applications 
Light and Shade and Their Applications 
Color and Its Applications 
Portable Lamps 
The Book of the Sky 
Lighting the Home 
Artificial Light; Its Influence upon Civilization 
The Language of Color 
The Lighting Art; Its Practice and Possibilities Representing a high standard of artificial lighting at the present time. (Courtesy of Mr. Ward Harrison, designer 
of the lighting equipment and installation.) LIGHT AND WORK 
A DISCUSSION OF QUALITY AND QUANTITY 
OF LIGHT IN RELATION TO EFFECTIVE 
VISION AND EFFICIENT WORK 
BY 
M. LUCKIESH 
DIRECTOR, LIGHTING RESEARCH LABORATORY 
NATIONAL LAMP WORKS OF GENERAL ELECTRIC CO. 
70 ILLUSTRATIONS 
D. VAN NOSTRAND COMPANY 
NEW YORK 
EIGHT WARREN STREET 
1924 Copyright, 1924, by 
D. Van Nostrand Company 
PRINTED IN THE UNITED STATES OF AMERICA 
THE PLIMPTON PRESS, NORWOOD, MASS., U. S. A. PREFACE 
DEVELOPMENTS in the production and the utiliza- 
tion of artificial light have been so great in recent 
years that for the first time in the history of mankind we 
are encouraged to hope for lighting conditions in the near 
future which approach the ideal. During the countless cen- 
turies preceding the present one, mankind had to be satisfied 
with mere light of very limited possibilities when daylight 
failed. It was less than a score of years ago that artificial 
light first showed promise of being a formidable competitor 
of natural light indoors. Now this promise may be realized 
because artificial light is not only a serious competitor of 
daylight on a cost basis but is far superior from the view- 
point of control. With the decreasing cost of artificial light 
it becomes increasingly practicable to simulate daylight in 
quality and to provide more desirable illumination intensi- 
ties. 
A score of years ago the discussions presented in these 
chapters would have been only of scientific interest. Now 
they are of very practical interest. The advent of promis- 
ing artificial light has stimulated investigation in the com- 
plex field of light, color, and vision. Researches in the 
laboratory and in actual practice have yielded valuable data 
which can be rendered of most practical value only by cor- 
relation and interpretation. Many questions still remain 
unanswered but it is ever so, for with more knowledge, more 
questions arise. Quality is the fundamental of light; 
quantity is the fundamental of lighting. These two fac- 
tors are constantly in mind throughout the discussions 
presented herewith. Design of lighting is not treated as 
such for the aim has been to present foundational material 
for the designer and user of lighting. 
v vi 
PREFACE 
Natural lighting outdoors is discussed as a powerful fac- 
tor in the environment of man during the period of evolu- 
tion and adaptation. Daylight is then discussed indoors. 
Artificial light is treated from the viewpoints of quality, 
quantity, and cost. Many fundamentals of vision and visual 
functions are treated from the viewpoint of efficient vision. 
The time element of vision is then introduced and this is 
followed by a discussion of the influence of intensity of 
illumination on production. An attempt is made to estab- 
lish the best range of illumination intensity and also the 
best quality of light for effective work as influenced by effi- 
cient vision. Some rather surprising conclusions are reached 
pertaining to the economic value of adequate and proper 
lighting and to the amount of artificial light that we can 
afford to use at the present time. It is hoped that the book 
will be helpful to lighting specialists, employers, workers, 
and all others interested in lighting in relation to the safety, 
the efficiency, the production, and the happiness of mankind 
in the field of work. 
February 12, 1924 
M. Luckiesh CONTENTS 
CHAPTER I 
PAGE 
LIGHT AND LIFE i 
Environment. Living Things. Primitive Man. Light. Evolution of 
the Eye. Radiant Energy or Radiation. Biological Effects of Radia- 
tion. Fundamental Lighting Conditions. What Good Lighting Ac- 
complishes. 
CHAPTER II 
DAYLIGHT OUTDOORS 20 
Sunlight. Effects of the Atmosphere. Skylight. Moonlight and Star- 
light. Nature's Landscapes. Fundamental Lighting Principles. 
CHAPTER HI 
DAYLIGHT INDOORS 47 
Daylight Openings. Intensity of Illumination. Reflection from Sur- 
roundings. Quality of Light. Hot-House Effect. Hygienic Value of 
Solar Radiation. Fading of Materials. Cost of Daylight. Windowless 
Buildings. Glasses for Daylight Openings. 
CHAPTER IV 
ARTIFICIAL LIGHT 71 
Fire. Primitive Light-Sources. Oil-Lamps. Gas Light-Sources. Electric 
Light-Sources. Electric Arc-Lamps. Mercury-Arcs. Electric Fila- 
ment-Lamps. Luminous Efficiency. Cost of Artificial Light. 
CHAPTER V 
ILLUMINANTS AND COLOR 86 
Spectral Character of Illuminants. Natural Illuminants. What Is Color? 
Ideal Illuminant for Color-Work. Accurate Artificial Daylight. Ap- 
proximate Artificial Daylight. 
CHAPTER VI 
QUALITY OF LIGHT AND THE HUMAN BEING 104 
Light of Natural Quality. Light of Unnatural Quality. 
CHAPTER VII 
QUALITY OF LIGHT n9 
The Sun as a Light-Source. Artificial Lighting and Decreasing Cost. 
Natural and Artificial Light Indoors. Influence of Daylighting on 
Artificial Lighting. 
CHAPTER VIII 
FUNDAMENTALS OF VISION 130 
The Eye. Illumination and Brightness of Test-Objects: Brightness Sen- 
sation. Visual Angle. Visual Acuity. Color Sensation. Simultaneous 
Contrast. Irradiation. The Pupil of the Eye. Adaptation of the 
Eye. The Purkinje Effect. Growth and Decay of Visual Sensations. 
External Factors in Vision. Binocular Vision. Shadows. High- 
lights. Backgrounds. 
vii viii 
CONTENTS 
CHAPTER IX 
PAGE 
SPEED OF VISION 176 
Speed of Discrimination of Very Fine Detail. Speed of Reading. Speed 
of Discrimination as Affected by "Confusion-Patterns." Simple Dem- 
onstrations. 
CHAPTER X 
LIGHTING AND PRODUCTION 199 
What May be Expected from Improved Lighting. Effect of Improved 
Lighting on Production. A Glimpse of Manufacturing Statistics. 
CHAPTER XI 
THE VALUE OF PROPER MAINTENANCE OF LIGHTING 
SYSTEMS 215 
Effect of Cleaning Glass of Daylight Equipment. Absorption of Light by 
Dirty Glass of Daylight Equipment. Depreciation of Artificial Lighting. 
Cost of Depreciation in Lighting. 
CHAPTER XII 
THE LIGHTING VALUE OF PAINT 225 
Influence of Ceiling and Walls. Reflecting Characteristics of Surfaces. 
Reflection Factors. Colors for Walls and Ceiling. Painted Machinery. 
Reflection of Invisible Radiation. 
CHAPTER XIII 
THE MOST EFFECTIVE INTENSITY OF ILLUMINATION .... 240 
Foot-candles, Reflection-Factors and Brightness. Plotting Data Per- 
taining to Visual Functions. A Resume of Data Bearing upon Illumina- 
tion Intensity and Vision. The Best Illumination Intensity Regardless 
of Cost. Low Contrast Requires More Light. Average Reflection- 
Factor. Factor of Safety in Lighting. 
CHAPTER XIV 
THE MOST ECONOMIC INTENSITY OF ILLUMINATION .... 261 
Cost of Artificial Lighting. Increased Production Can Purchase More 
Light. Form of Curves Representing Increase in Production. 
CHAPTER XV 
VISIBILITY AND SAFETY 279 
Glare. Brightness of Light-Sources and Lighting Equipment. Eye- 
Strain. Eye-Fatigue. Visibility. Safety and Accidents. Traffic. 
Traffic Signals. 
TABLES 
I. Some properties of radiation 13 
II. Percentages of direct sunlight and of skylight in the total light reach- 
ing a horizontal surface at noon for various sky-conditions .... 31 
HI. Average brightness of the sky under various conditions 32 
IV. Reflection-factors of earth areas 38 
V. Transmission-factors of typical glasses 69 
VI. Approximate temperatures and luminous efficiencies of a few light- 
sources 92 CONTENTS 
ix 
PAGE 
VII. Natural and artificial illuminants arranged in relation to each other 
along the natural scale of spectral character from flames to blue 
sky 96 
VIII. Results of an investigation of the preferences of a group of observers 
for two illuminants, namely, that from regular Mazda C lamps 
and that from regular daylight lamps 109 
IX. Showing the preference of a mixture of approximate artificial daylight 
with natural daylight over a mixture of light from equal wattages 
of tungsten gas-filled lamps with the natural daylightno 
X. Choice of intensities of illumination of lights from mercury-arc and 
gas-filled tungsten lamps 117 
XI. The relation of size of objects and their distance from the eye for small 
visual angles 147 
XII. Showing the relation of intensity of illumination to factory production 207 
XIII. Summary of statistics of manufactures for the United States .... 211 
XIV. Statistics of manufactures for cities of 10,000 and over and for cities 
of 100,000 and over 211 
XV. Results of cleaning glasses in factory day lighting equipment 219 
XVI. Showing the influence of the reflection-factor of the walls of a spherical 
room upon the illumination of the walls. An unshaded light-source 
is at the center of the sphere of such intensity that it illuminates 
the sphere with uniform direct light to an intensity of one foot- 
candle 227 
XVII. Showing relative foot-candles on work-plane from a given number of 
lamps used in three different lighting systems for various reflection- 
factors of walls and ceiling and for three sizes of rooms 229 
XVIII. Reflection-factors (in per cent) of common dry powdered pigments 
for various illuminants 235 
XIX. Data pertaining to the range of brightness (and illumination inten- 
sities) for efficient vision. These data are taken chiefly from the 
illustrations indicated 249 
XX. Visual sensitometric data 251 
XXI. Summary of data from Table XIX 254 
XXII. Costs of installation and of operation per year (2500 hours) of a lighting 
system for a factory 100 ft. by 200 ft 263 
XXIII. Showing how much light can be purchased by the increase in produc- 
tion resulting from a given increase in intensity of illumination. 
The lighting system and the number of workers remaining constant 
but the illumination intensity and wage-scale being changed. ... 270 
XXIV. Approximate brightness in candles per square inch 282 LIST OF ILLUSTRATIONS 
FIGURE PAGE 
1. Various spectral regions and properties of radiation of different wave- 
lengths 12 
2. Variation of skylight, sunlight and total daylight on a horizontal surface 
throughout a typical clear day in midsummer 22 
3. The maximum possible duration of sunlight in midsummer and in mid- 
winter for different latitudes in the northern hemisphere 23 
4. Showing the variations in the ratio of the intensity of skylight illumination 
to that of total daylight illumination on a horizontal surface through- 
out four typical days. These different days may be described respect- 
ively as dense, slightly broken clouds; thinly overcast sky; clear day 
with few clouds; and clear day 28 
5. Variations in the brightness of patches of sky about 20 degrees above the 
horizon in the north, east, south, and west respectively, throughout 
most of a clear day in August at 40 deg. north latitude 34 
6. Showing the effective sky-area at a surface at some distance from a window 
and also showing the general direction of light from the sky-area to the 
surface 54 
7. Four typical crystal glasses having different surface characteristics as 
indicated by their respective cross-sections 66 
8. Showing the actual transmission of light through the four typical crystal 
glasses having surface contours as shown in Figure 7 67 
9. Showing the approximate percentages of light reflected by the first and 
second surfaces of clear plane glass at three different angles of incident 
light. The net transmissions are shown approximately, assuming no 
absorption of light within the glass 67 
10. Showing the percentage of incident light reflected by one and two surfaces 
of clear clean plane glass for various angles of incidence 68 
11. Illustrating the refraction of light by a " prism " glass 69 
12. Illustrating total reflection within glass prisms 70 
13. Average luminous efficiencies in lumens per watt of electric filament lamps 
in use each year from 1905 to 1923 80 
14. Luminous efficiencies of various electric filament lamps 81 
15. Relative costs of artificial light during various periods of the past century. 82 
16. The decreasing cost of electric lighting and the increasing cost of food for 
the twenty years preceding the World War. The war period is left 
out of consideration, owing to the abnormal condition; however, it 
is interesting to note that the cost of electric lighting since 1914 has 
steadily decreased while the cost of food has greatly increased .... 83 
17. The approximate average lumens of light used in the United States per 
inhabitant each year from 1912 to 1923 85 
18. The spectral character (spectral distribution of radiant energy) of the 
visible radiation reaching the earth from the sun at different altitudes 
or periods of the day. It is seen that the sun appears reddish toward 
sunset owing to the predominance of yellow, orange, and red rays . . 88 
19. The spectral distribution of visible radiation in various illuminants .... 90 
20. Showing the influence of cloudiness on the amount of electric current used 
for artificial lighting. The dotted line represents the daylight illumina- 
xi xii 
LIST OF ILLUSTRATIONS 
FIGURE PAGE 
tion intensity outdoors (right-hand scale) on a horizontal surface ex- 
posed to the direct sunlight and the entire sky on July 29. The two 
other curves are total current consumption for a cloudy day (July 29) 
and a clear day (July 25) respectively 127 
21. Illustrating the various parts of the eye by a vertical cross-section along 
the optical axis 131 
22. Showing the minimum brightness that is perceptible at various bright- 
nesses corresponding to those produced by a wide range of intensities 
of illumination (foot-candles) of a so-called white surface of 80 per cent 
reflection-factor. The horizontal scale has been expressed in foot- 
candles on a colorless surface which diffusely reflected 80 per cent of the 
incident light as nearly as possible from the original data presented by 
Kbnig. The results for white light and also spectral blue and red lights 
are shown 139 
23. Showing the sensitivity of the eye to brightness-difference throughout a 
range of brightnesses corresponding to illumination intensities up to 
25-foot candles on a so-called white surface of 80 per cent reflection- 
factor. The upper horizontal scale is the same as the lower one when 
both scales are reduced to brightness values because the upper scale 
is for foot-candles on a colorless surface that reflects only 8 per cent of 
the incident light. By sensitivity is meant the ability of the visual 
sense to see brightness-difference 141 
24. Showing the per cent increase in the sensitivity of the eye to brightness- 
difference for surfaces receiving full illumination (solid line) represented 
by the foot-candle scales and for a shadow (broken line) receiving only 
10 per cent of this full illumination. The relative sensitivity is repre- 
sented by 100 at the brightness of a colorless surface of 80 per cent re- 
flection-factor under one foot-candle of illumination (or a surface of 
8 per cent reflection-factor under 10 foot-candles) and for a shadow 
(broken line) receiving only 10 per cent of the full illumination . . . 143 
25. Showing the relations of size of an object and its distance from the eye for 
small visual angles. The straight line represents the sizes of objects in 
hundredths of an inch (right-hand scale) at a distance of 14 inches 
from the eye for small visual angles. The curve shows the ratio of the 
distance of an object from the eye to the size of the object in the same 
units of length (left-hand scale) for small visual angles 146 
26. The solid line represents the general relation between visual acuity and 
illumination intensity for a normal eye, the test-object consisting of 
black details on a so-called white background of 80 per cent reflection- 
factor (lower scale). The upper scale applies to a dark gray back- 
ground of 8 per cent reflection-factor. The dotted line shows the 
general relation of the results obtained with a slightly astigmatic eye . 148 
27. The solid line shows the per cent increase in visual acuity with brightness 
of background against which the black details are viewed. The foot- 
candle scales apply to surfaces of 80 per cent and 8 per cent reflection- 
factors respectively. The relative visual acuity is represented by 100 
for a brightness of background corresponding to one foot-candle on a 
colorless surface of 80 per cent reflection-factor (or 10 foot-candles on a 
colorless surface of 8 per cent reflection-factor). The broken line shows 
the per cent increase of visual acuity in a shadow receiving only 10 
per cent of the full illumination represented by the foot-candle scales . 150 
28. Showing relative visual acuity for various portions of the retina. Zero de- 
grees is at the center of the fovea of the retina. The dotted line repre- 
sents the relative visual acuity at ordinary brightnesses of high-intensity LIST OF ILLUSTRATIONS 
xiii 
FIGURE PAGE 
illumination. The solid line represents visual acuity for very low bright- 
nesses (low intensities) 152 
29. The effect of pupillary aperture on visual acuity. The full line A represents 
results obtained with various apertures and a test-object of constant 
brightness. The broken line B represents results obtained with various 
pupils but with the brightness of the test-object altered in each case to 
keep the retinal image of constant brightness. The brightness of the 
test-object was the same for the 1 mm. pupil in the two cases .... 158 
30. Showing the size of the pupil after 15 minutes' exposure to various bright- 
nesses from zero to 2000 millilamberts (about 4 candles per sq. in.) . . 161 
31. Showing the rate of opening the pupil after a brightness of 100 millilamberts 
has been replaced by darkness; and the rate of closing after darkness 
has been replaced by the stimulus of 100 millilamberts 162 
32. The increase in sensitivity of the eye with time of exposure to darkness as 
determined by the minimum brightness perceptible as the time of 
adaptation to darkness increased 163 
33. Illustrating the overshooting of brightness sensation. The figures are foot- 
candles and the brightnesses are those produced by these illumination 
intensities on a so-called white surface of 80 per cent reflection-factor 167 
34. Showing the duration of after-images from exposures of various lengths to 
light-sources of different brightnesses 169 
35. The frequency at which flicker disappears for surfaces of various reflection- 
factors illuminated to various intensities as indicated by the horizontal 
scale. The flicker was caused by abrupt changes from bright to dark, 
the periods of bright to dark being equal 171 
36. The international test-object for visual acuity. The dimension of the 
opening a determines the visual angle when distance from the eye is 
taken into account. The subject not only must see the discontinuity 
of the ring but must state in what position it is. Several sizes and posi- 
tions are reproduced herewith so that the reader may compare them 
at a distance at which the smallest is not seen as a split ring 180 
37. Showing the speed of discrimination as affected by intensity of illumination 
(brightness of background) for black test-objects of four different sizes 
(visual angles 1.15, 1.7, 2.49 and 3.45 minutes). The lower scale repre- 
sents the illumination of a colorless background of 80 per cent reflec- 
tion-factor against which the test-object was viewed. If the background 
were a dark gray of 8 per cent reflection-factor, the foot-candle 
must be multiplied by 10. Thus the upper scale fits the data equally 
well assuming that the test-object wras truly black on the background 
of 8 per cent reflection-factor 181 
38. Data in Fig. 37 for the smallest and largest test-objects are plotted in terms 
of per cent increase in speed of discrimination for the same range of 
foot-candles on the two backgrounds (so-called white and dark gray) . 182 
39. Showing the relation between visual angle and speed of discrimination 
for three intensities of illumination on the black-white test-object. If the 
text-object were truly black and were seen against a background 
of about 8 per cent reflection-factor these foot-candle values would be 
ten times greater 183 
40. Showing the speed of discrimination of a black dot (2.43 minutes visual 
angle on a so-called white background (80 per cent reflection-factor) 
for a range of illumination up to 150 foot-candles. The upper scale 
is for a truly black dot on a dark gray background of 8 per cent reflection- 
factor 186 
41. Apparatus used for speed of reading tests. Old English print revolved on xiv 
LIST OF ILLUSTRATIONS 
FIGURE PAGE 
the drum C under the slit E which was viewed from H with the head 
resting on K. The light source was G in the shield F and the intensity 
was controlled without change in quality. D was a screen of the same 
general brightness of the printed matter at all times 189 
42. Effect of intensity of illumination on the speed of reading. A is for a low- 
contrast test-object (black print on a background of 23 per cent re- 
flection-factor). B is for the usual condition of black print on a white 
paper of 80 per cent reflection-factor 191 
43. Relation of brightness of background (the blank page) and speed of reading. 
(Data from Fig. 42.) A is for low contrast (black print on gray paper 
of 23 per cent reflection-factor) and B is for high contrast (black print 
on white paper of 80 per cent reflection-factor) 192 
44. The test-object b is preceded by a and followed by c. These confusion 
fields a and c have a marked influence upon the time required to dis- 
tinguish b 193 
45. Curve A is the relation of the speed of discrimination of the test-object 
b, in Fig. 44, to the intensity of illumination of the white background 
of 80 per cent reflection-factor. Curve B shows the results when the 
test-object b is preceded by a and followed by c, Fig. 44. The upper 
scale represents the relation between speed of discrimination and foot- 
candles if the test-object had been truly black on a dark gray back- 
ground of 80 per cent reflection-factor 194 
46. A simple device for demonstrating the effect of intensity of illumination 
on the time required to distinguish an object, a is a slot in the front of 
an envelope and a letter E is on the back, c. The slide b is inserted and 
two rubber bands R are slipped over as shown in the assembly. By 
pulling the slide out against the bands different exposures of the E 
can be made, d shows the wedge-shaped slot in which several E's can 
be exposed for different periods simultaneously 197 
47. Showing the average number of pieces of material inspected per person 
per hour for each of ten consecutive weeks; also the corresponding in- 
tensities of illumination for the various weeks. During the first two 
weeks the old lighting system (half daylight) was in operation. During 
the remaining eight weeks the new system of artificial lighting was 
operated 205 
48. A graphic illustration of the per cent increase in production and of the 
cost of lighting in terms of the payroll for the various intensities of 
illumination. The production increases are averaged from Fig. 47. . . 206 
49. Showing the relation of illumination intensity to the increase in production 
and to the cost of light in terms of the payroll 208 
50. Showing the billions of dollars of value which could be created without 
cost each year and also saving in labor cost for various increases in 
production due to improved lighting throughout the manufacturing 
industries of the United States. These values are based on the 1919 
census of manufactures 213 
51. Chart showing the importance of prompt renewal of burned out lamps and 
systematic cleaning of the lighting equipment 220 
52. Common reflection characteristic of surfaces: a, a regular or specular 
reflection; b, perfectly diffuse reflection; c, spread reflection 230 
53. Showing the amount of regular and of diffuse reflection of various surfaces 
which exhibit a combination of these two types of reflection. The 
regular reflection is determined for perpendicularly incident light ... 231 
54. Showing the illumination intensities in foot-candles for surfaces of various 
diffuse reflection-factors necessary to produce a brightness of one millilam- LIST OF ILLUSTRATIONS 
FIGURE PAGE 
bert. The left-hand scale should be used with the lower scale for all values 
of reflection-factor greater than io per cent. For smaller values the 
upper and right-hand scales should be used 242 
55. The equation of the curve is y = log x. y is represented by the vertical 
scale on the left for corresponding values of x on the lower scale for 
a certain range of values (from x = o to x = 100). The identical diagram 
serves to represent another range of values (x = o to x = 0.1) as seen 
by applying the scale to the right and the upper scale 245 
56. Showing the influence of intensity of illumination (or brightness of back- 
ground) on the speed of discrimination of a test-object consisting of 
bars 3 mm. wide separated by intervals of the same width and viewed 
at 19.7 feet. The test-object was similar to b, Fig. 44, and it was pre- 
ceded and followed by confusion patterns similar to a and b (Fig. 44) 
respectively. The high contrast (100:5) consisted of the usual "black" 
on "white." The low contrast (100:73) consisted of light gray bars 
on "white." The data are the averages of 21 observers 255 
57. Data from an investigation similar to that whose results are plotted in 
Fig. 56. Same observers, apparatus, etc., except that the bars of test- 
object (and intervals between) were 2 mm. wide. Note the relatively 
high intensities of illumination required to attain the same speeds of 
discrimination as in Fig. 56 256 
58. Cost of operation per year (2500 hours) of a lighting system yielding a 
uniform illumination on the work-plane for various intensities. The 
room is 100 ft. x 200 ft. RLM dome porcelain-enamelled reflectors, 
containing bowl-enamelled tungsten lamps, are spaced uniformly on 
ro-foot centers 264 
59. Fixed charges, cost of electric energy at 3 cents per kilowatt hour, and 
total annual cost of operation for various uniform intensities of illumi- 
nation on the work-plane 264 
60. Lamp renewals, depreciation, interest and insurance for various 
uniform intensities of illumination on the work-plant 265 
61. The operating costs of the lighting reduced to cents per foot-candle per 
square foot per year. Electric energy being such a large part of the 
total, the results are given for three rates, namely 1 cent, 3 cents, and 
5 cents per kw. hr 266 
62. Illustrating how much light can be purchased for actual increases in pro- 
duction. Data represented by A, C, and E are same as in Fig. 48. 
The line BD extended both ways represents cost of light in terms of 
payroll (average wage 30 cents per hour). The line GH represents cost 
of light in terms of a doubled payroll (average wage 60 cents per hour). 
In this case, the production under the old lighting is taken as a basis. 
By projecting horizontally from E for example, the point F indicates 
the foot-candles (103) of the new lighting that could be purchased by 
the increase in production 267 
63. Similiar to Fig. 62 excepting that the production under 6 foot-candles 
from the new lighting system (Fig. 48) is taken as a basis. This elimi- 
nates the matter of different systems of lighting. The cost of light in 
terms of the payroll is represented for several wage scales, 30 cents 
per hour being the actual one at the factory where Hess and Harrison 
obtained the data 269 
64. The simplest equations which fairly well represent the data presented in 
certain illustrations in Chapter IX have been determined. They are 
represented by the curves. To show how well they represent the actual 
data, some of the latter are indicated in circles 273 
xv xvi 
LIST OF ILLUSTRATIONS 
FIGURE PAGE 
65. An equation was determined for the data on increase in production Fig. 48 
and is represented by the curve EAD. The cost of light in per cent of 
payroll (average wage 30 cents per hour) is represented by BD extended 
both ways. EAD intersects BD at 160 foot-candles. The cost of light 
in per cent of payroll if average wage were 60 cents per hour is also 
represented 276 
Representing a high standard of art lighting at the present time . . . Frontispiece 
Before and after cleaning the glass skylights Facing page 47 
Representing the result of indifference to lighting " 199 
Reflections-factors of colored surfaces " 235 LIGHT AND WORK 
Chapter I 
LIGHT AND LIFE 
EVER since primitive man began to think of his own 
potentiality and began to resist the whims of his en- 
vironment, he has been building for himself and his posterity 
an artificial world. He is incessantly engaged in un- 
ravelling the mysteries of Nature and he weaves the knowl- 
edge gained into a fabric for his own specific use. The 
natural environment can be no more than a compromise, for 
into it is woven a very complex life which, doubtless, is 
as well adapted to survive as is possible in such a com- 
plexity. Man cannot escape the fact that he is still Nature's 
being, subject to Nature's laws. However, with knowledge 
of these laws and of the imprints which natural environment 
have left upon him, he should be able to make a better 
environment for himself alone than Nature provided for him 
as merely a detail of a complex whole. 
It is not the intention to attempt to discuss whether or 
not mankind is progressing toward salvation or downfall 
in building to some extent an independent artificial world. 
Any one of the great fundamental elements of the natural 
world provides a problem complex enough. Light and its 
associates - color, lighting, and vision - and their in- 
fluence upon the economic activities of civilized peoples are 
of primary importance in this artificial world now under 
construction. Man evolved under an environment in which 
very adequate lighting by means of light of certain char- 
acteristics was, and still is, a factor of equal importance 
with " air, earth, and water." By coming indoors he 
created certain restrictions on daylighting. Furthermore, 
1 2 
LIGHT AND WORK 
we find man, product of an environment of adequate light 
of certain qualities, forced to begin at the other end of the 
scale in the development of artificial lighting. Whether 
or not man is tempting fate in building this elaborate arti- 
ficial world is a question that can not be answered until this 
structure is more nearly complete and has a record of opera- 
tion. The pertinent fact is that he is building it and desires 
artificial light for use where Nature's light cannot penetrate 
and where Nature's light daily fails - a failing to which 
life adapted itself as a period of rest. 
However, leaving all that aside, we find man in darkness 
" making light." We find him developing an art of arti- 
ficial lighting, beginning with darkness which is so far from 
the tremendous intensities of lighting in which he and his 
physiological and mental processes evolved. We may pic- 
ture man in ecstacy over his successes in making light. In 
comparison with natural lighting, he has done little more 
up to the present time than dissipate darkness in the vicinity 
of his light-sources. It is as if he had lighted a match in a 
huge dark room. But being able to see here and there, he 
goes on with his daily work serene in the conviction that 
he has adequate lighting of the proper qualities. He little 
realizes that, so far, he has merely dispelled darkness here 
and there by means of light of more or less accidental 
qualities. In general, he does not know that the advantages 
of lighting increase in proportion to the progress of arti- 
ficial lighting up the scale from darkness toward the 
conditions of natural lighting. He is not generally conscious 
of the fact that, if artificial lighting is founded upon the 
fundamental principles of natural lighting and includes im- 
provements upon natural lighting where such are obviously 
safe and desirable, his activities will be the most efficient, 
productive and enjoyable. 
The economic value of the results in which visual proc- 
esses play an essential and important part is almost iden- LIGHT AND LIFE 
3 
tical to man's entire achievements because he is so 
dependent upon vision. It follows, not only logically but 
from experiments and general experience, that the eco- 
nomic advantages which would result from raising the 
standard of artificial lighting to a level even now practicable, 
would be very great. It is necessary to realize that mere 
light and haphazard lighting eliminate darkness but do 
not provide the best economic lighting conditions which are 
now possible even in this early stage of mankind's progress 
toward independence or improvement upon natural condi- 
tions. It is the aim of this book to analyze natural light- 
ing, to point out its important fundamentals, to suggest im- 
provements, to reveal the relations between lighting and 
vision, and to show the economic results of adequate and 
proper lighting in increasing the safety, efficiency and hap- 
piness of mankind. To those who must appraise progress 
in terms of gold, the promise is given that improved lighting 
in this country would be worth billions of dollars annually 
for it can create this value in excess of its cost. 
Environment 
No adequate conception of the powers of light and of 
the possibilities of modern light can be achieved without 
a study of the environment under which the human race 
evolved. One of the most important environmental factors 
throughout the eons of evolution of life was light, or more 
broadly, solar radiation. The quantity and the quality of 
sunlight are fundamental characteristics to which the visual 
sense-organ became adjusted throughout these ages of 
development. The direction of the dominant light was so 
powerfully influential during the development of the human 
body that the physiognomy was moulded to best suit it. 
The eyebrows were projected forward to screen the eyes 
somewhat from the dominant light-sources - the sun and 
the sky. 4 
LIGHT AND WORK 
The distribution of brightnesses outdoors where the 
human race developed, has left its imprint. A " natural " 
distribution of brightness to us in general has relatively 
lower brightnesses below the horizontal as compared with 
the brightnesses in the upper portions of the visual field. 
In other words, the sky is the brightest of the large areas 
and this is usually the most pleasing to us. On the other 
hand, the sun is a source of direct light which has also 
played its important role with lasting effect upon the human 
being. All these factors and many others discussed in 
later chapters, have moulded us physically, physiologically 
and psychologically. And these influences of the continued 
adaptation of ages cannot be wiped out in a short time. 
Perhaps the psychological imprints of these environmental 
factors of natural lighting are most susceptible to alteration. 
Certainly the physiological factors of vision cannot be gen- 
erally altered in a short time. It seems reasonable to be- 
lieve that they would require nearly as many centuries of 
adaptation to another environment to alter them as it would 
to relocate protective eyebrows by adaptation to a radically 
different location of the dominant light-sources. The re- 
ceptors in the human eye have been adapted to certain wave- 
lengths of radiation, to certain intensities of illumination, 
and to various other factors which are discussed in these 
chapters. 
These great facts of environment which have made 
us what we are and our sense-organs what they are, and 
which have established our psychological attitude, cannot 
be ignored if we are to get the best out of lighting and of 
the human being. Still they are almost universally ignored. 
What thought does the factory manager, and others who 
employ labor for its output, give to these matters? How 
oblivious most persons are to the importance of light and 
particularly of proper and adequate lighting. Even many 
of those dealing with lighting directly, lose sight of the im- LIGHT AND LIFE 
5 
portance of lighting conditions similar to those Nature pro- 
vided. It is easy to forget that what is, was not always. 
It is easy to be so intent upon today that sight is lost of the 
fact that artificial light made its advent only yestercentury. 
In this highly artificial age when conditions are changing 
rapidly we should be certain that change means progress. 
One of the details of this high-speed age is our dependence 
upon artificial light. Therefore, every practicable advan- 
tage should be grasped and this cannot be done without a 
breadth of view obtained by a study of natural lighting. Cer- 
tainly by lighting a match or a glaring lamp we have not 
overcome all the handicaps of darkness. Certainly this act 
does not span the gap between darkness and outdoor day- 
lighting conditions. For this reason these early chapters 
are designed to give the reader a view of lighting conditions 
to which our eyes, bodies, and minds were adjusted through- 
out the ages. But first let us consider further, though briefly, 
some important fundamentals. 
Living Things 
Light is as essential to most kinds of life as it is to vision. 
It is as indispensable to nearly all plants and animals as 
it is to practically all human activities. It is one of the 
foundational stones of the universe. Mere accident has not 
made it essential to life and to civilized progress. Its over- 
whelming and universal importance indicates that light is 
one of the few materials from which the mechanism of the 
universe is composed and energized. Certainly it plays a 
fundamental role in the evolution of life. If we look back- 
ward to that distant period which exists in our imagination 
when we see its history written in geology, it may occur to 
us to ask why the evolution of life was such as to make light 
essential to it. In that early period when the earth was 
still steaming, darkness had just the same opportunity as 
light. But darkness succumbed to light as a potent power. 6 
LIGHT AND WORK 
At best we may state that darkness is the period of general 
rest. 
In the dark cellar or cave what life is found, is so different 
from that found in the light. In those unillumined places 
plant-life is ghostly - at best, delicate and unreal. The 
animal life is usually repulsive - at best, unattractive to 
us. At night most birds are asleep. Some of the more ex- 
uberant or optimistic ones, such as the song-sparrow and the 
mocking bird, occasionally burst into song after dark but 
usually the night is given over to bats and owls. Predatory 
animals are abroad after dark, but this is a matter of neces- 
sity for their preying habits require the shelter of darkness. 
Only those who are weak or bent upon mischief operate be- 
hind the screen of darkness. Of course, we can learn to be 
useful within a limited sphere, of activities without light. In 
proof of this behold the blind person; but his uncertain step 
as he feels his way is pitiful evidence of the absence of the 
security and certainty which light provides. 
The life-giving value of light is seen on every hand. 
Plants require light for normal growth and, in general, they 
require much more light for healthy growth than animals. 
Most of them require the intensities of daylight hundreds 
of times greater than those commonly encountered in arti- 
ficial light. Man and many animals seem to live in comfort 
and health under intensities of illumination which are inade- 
quate for most plants. Even in a well-lighted room plants 
grow toward the window for they want even more light. 
They are used to the outdoor intensities and most of them 
are sickly if their lighting is restricted to the extent that 
man has restricted his lighting. 
Primitive Man 
What effect has the donning of clothing and the coming 
indoors had upon mankind? Has it been good or bad? Cer- 
tainly the ancestors of civilized man were at one time LIGHT AND LIFE 
7 
adapted to the great intensities of illumination outdoors 
in the daytime. Primitive races are now living under these 
conditions outdoors. Would they survive a sudden trans- 
planting to the indoor environment of civilized man? The 
human race came indoors gradually and this gradual change 
was accompanied by knowledge which substituted some- 
what for the blessings of the sun. For example, civilized 
man gives some consideration to sanitation and personal 
hygiene. It is true that he has not at any time in his 
history been too exercised in these directions, but at 
least he has given them much more consideration 
owing to his intelligence than the savage would give 
them if he suddenly came indoors. With the sun no 
longer acting directly upon his body or wound or scanty 
apparel, the savage would be in grave danger from germs. 
Lacking in knowledge and inclination this savage race would 
likely decay and even perish if it withdrew from the sun 
to the extent that most civilized persons have. 
But knowledge cannot overcome in a short time the re- 
sults of ages of environment. It cannot alter physiological 
and psychological laws and processes that are the products 
of long adaptation. The eye cannot function best under a 
low intensity of illumination when it has been developed 
for serious activity under the great intensities outdoors. 
Continued vision is not likely to be as satisfactory or free 
from fatigue and harmful results under conditions radically 
different from the general conditions outdoors. Illuminants 
differing radically from daylight, and particularly from sun- 
light, may be logically subjected to close inquiry if they are 
to be used for long periods for serious visual work. Psycho- 
logical effects of radically different lighting conditions are also 
proper subjects of investigation. Man, in building his arti- 
ficial world, wrests secrets from Nature and uses this knowl- 
edge in his struggle for independence. He has a right to 
improve upon Nature and he often does so. However, 8 
LIGHT AND WORK 
Nature's way is the safe way until man is sure that he has 
equalled or improved upon it. 
Throughout the ages during which man knew nothing of 
the laws of the physical world, he was purely a product of 
environment - not only of his own environment but of that 
of his ancestors. In this condition he was also evolving 
helplessly and unconsciously into a future over which he 
could exercise no influence without knowledge. In the 
earlier ages, of course, he noticed that a rock was heavy 
and a feather was not but he more generally learned facts 
pertaining to his own immediate material needs. His crude 
learning was a matter of experience. His instincts were the 
result of the experience of ancestors. But his knowledge 
of natural phenomena and their laws and relations remained 
meagre indeed until recent centuries when serious inquiry 
by experimentation began. As far as the human race is 
concerned, experimental research has just begun and the 
job to be done is of great magnitude. Therefore, it is not 
surprising that we often meet questions that we cannot 
answer as yet. Naturally we know more of physical phe- 
nomena than of the physiological and psychological phase 
of life. This is particularly true of light and of radiation 
in general. However, it is true now, and always will be, 
that practice lags behind knowledge. 
Light 
In respect to light we have the physical energy which we 
speak of as radiation or radiant energy. For example, 
such energy, emitted by the sun, may be termed solar radia- 
tion. It is emitted in all directions in space and some of it 
is intercepted by the relatively small body which we call 
the Earth. Here this radiation performs many functions. 
Some of it is absorbed and changed into heat energy, 
thereby warming the absorbing objects. Some of this 
radiation kills germs. Some of it produces chemical changes 
resulting in the growth of plants. Some of it is received by LIGHT AND LIFE 
9 
the eye resulting in a sensation of light. There are many 
interesting effects of solar radiation but here we are par- 
ticularly concerned with the human being and the effect of 
light and accompanying radiations. (See Fig. i) 
Solar radiation consists of radiant energy of many wave- 
lengths. It can be separated according to wavelength by 
means of prisms and other devices. On doing so we find 
that the eye is sensitive to only a certain range of these 
wavelengths. It is a receiving station and the sun is a 
sending station. This radiant energy is really electro- 
magnetic radiation differing only in wavelength from that 
used in wireless telegraphy or " radio." The eye responds 
to only a certain narrow range of wavelengths just as the 
" radio " receiving station. These radiations of various 
wavelengths to which our visual sense-organ is sensitive pro- 
duce sensations of color.1 Beginning with the shortest wave- 
length of radiation to which the eye is " tuned," the result 
is a sensation of violet. With increasing wavelength there 
is a progressive change in the color-sensation which passes 
through blue, green, yellow, orange, and red. These are 
merely the most distinctive colors. There is no dividing 
line between them. Theoretically there are as many dif- 
ferent hues as there are different wavelengths within the 
range of sensibility of the eye. Practically, we can distin- 
guish only about 125 distinctly different hues. If a normal 
eye receives all these radiations at the same time and in 
the same relative amounts emitted by the sun (as seen at 
noon on a clear day) the resulting sensation is a combina- 
tion of all the possible individual ones. This resultant sen- 
sation is approximately white. This is an extremely im- 
portant matter as later chapters will show. 
Evolution of the Eye 
Let us pause a moment to consider a very interesting fact. 
We have just discussed the visible spectrum or the range 
of wavelengths of radiation which our visual sense-organ 10 
LIGHT AND WORK 
converts into sensations of light or color. If we measure the 
relative amounts of energy of these various wavelengths 
reaching the earth from the sun at midday, we find that 
the maximum amount is that of wavelengths which produce 
the sensation of yellow-green. If we measure the relative 
sensibilities of the eye for radiations of different wave- 
lengths we find that it is most sensitive to that which pro- 
duces the sensation of yellow-green. In other words, if 
we produce a normal visible spectrum in which equal 
amounts of energy of all visible wavelengths are present, we 
find that it is brightest in the yellow-green portion. This 
is not strange for the human eye is the result of countless 
ages of evolution under an environment in which sunlight 
is a prominent factor. Animal life that followed other 
courses of evolution which resulted in different environ- 
ments shows approximately different characteristics. 
Water of considerable depth is bluish green in color. There- 
fore, only light of this color is transmitted to this depth. 
Fish living at these depths need not have eyes sensitive to 
all the radiations that the human eye can see and it is in- 
teresting to note that their color-vision is correspondingly 
restricted. None, or at best only the crudest eye, is found in 
life beyond the reach of solar radiation, such as in caves and 
at the bottom of the sea. It is against a fortress of such 
evidence that certain puny men with strange minds which 
do not seek or recognize truth, are training their antiquated 
guns loaded with bullets of blind prejudice and powder of 
ignorance. 
Radiant Energy or Radiation 
The radiations to which the eye responds and converts 
into sensations of brightnesses and hues (or colors), may be 
grouped in the term visible radiation and their visual effects, 
in the term light. These radiations, when separated, are 
seen as a series of colors as in the rainbow and this series LIGHT AND LIFE 
11 
of colors is termed the visible spectrum. Accompanying the 
visible radiation from the sun and other sources of radiant 
energy, are invisible radiations of various wavelengths. The 
region of longer wavelengths than that which produces in 
our visual sense-organ the sensation of red, is termed infra- 
red radiation. The region of shorter wavelengths than that 
causing the sensation of violet is termed ultraviolet 
radiation.2 
It broadens and, at the same time, simplifies our view 
of radiant energy from the sun and other sources, to consider 
that the eye is just one of many " receiving stations " which, 
in general, differ in range of wavelengths to which they are 
sensitive. For example, the ordinary photographic emul- 
sion is sensitive to wavelengths which we see as blue and 
violet light and also to a great range of wavelengths in the 
ultraviolet region. Plants absorb radiations of various wave- 
lengths. There are many photochemical reactions each of 
which is caused by radiations of certain wavelengths 
peculiar to it. In all cases where radiant energy is absorbed 
some result is achieved such as chemical reaction or a rise 
in temperature. Perhaps it should be conversely stated that 
any effect of radiation is due to an absorption of it and some 
kind of transformation. 
These invisible radiations are not of primary importance 
from the viewpoint of light, but they must be considered 
in connection with Nature's lighting. They are a part of 
the environment under which life evolved and inasmuch 
as this environment is foundational with respect to this 
book they cannot be ignored. Furthermore, these radia- 
tions accompany the visible radiation from artificial sources; 
from the viewpoint of light-production, they represent 
waste and, therefore, inefficiency. On the other hand, they 
are being turned to good use in many chemical reactions, 
in therapy, in sterilization, etc.2 Our chief interest in them 
is in the safe assumption that the human being is (or was 12 
LIGHT AND WORK 
before he came indoors) adapted to those found in solar 
radiation. We know less about their applications than we 
do about visible radiation or light, but we are learning 
rapidly. Here, as in other fields, we are improving upon 
Nature by producing specific radiations for specific purposes. 
Transmission of different depths of 
water for short-wave infra red radiation 
Absorbed bvafew inches of water 
Partially absorbed by water-vapor 
Fig. i. Various spectral regions and properties of radiation of different wavelengths. 
/4-00 
~O024in> 
Q3 in. 
Transmitted by ordinary clear glasses 
Short-wave infra-red 
/ZOO 
WAVELENGTH OF RADIATION 
Transmitted by cornea of eve 
Transmitted to retina of eye 
1000 
Solar radiation 
900 
Percent 
Transmission 
800 
80 
40 
•60 
20 
o 
Transmitted by few mchesofwater 
Hue Sensations 
700 
Yellow 
Orange 
Red 
600 
i 
Visible 
.Green 
Relative 
Luminosities 
500 
Blue 
VioJet 
Near Ultraviolet 
400 
-JL_ 
300 
i 
Irritating toeye 
Absorbed by eye 
Middle Ultraviolet 
zoo 
Ultra violet 
Absorbed by 
Ozone 
Germicidal 
Extreme Ultraviolet 
loo 
Absorbed by 
most media 
Radium Pdy^poen^enpd/J 
0 
In Fig. i, are shown graphically, on a true scale of wave- 
lengths, the ultraviolet, visible, and near infra-red regions; 
the short-wave limit of transmission of quartz, water, and LIGHT AND LIFE 
13 
ordinary glasses; the short-wave limit of solar radiation 
reaching the earth; the long-wave limit of germicidal action; 
the relative visibility of visible radiations of various wave- 
lengths and the sensations of hue which they give rise to. 
These limits are only approximate because, in general, they 
cannot be exactly defined excepting for a specific case. The 
infra-red region extends much further than shown and this 
is indicated by leaving the diagram unclosed at the right. 
However, the details of infra-red radiations of larger wave- 
lengths are of less interest than the radiations of the 
wavelengths shown. It will also be noted that no such 
terms as " actinic " or " chemical " rays are used. These 
are vague terms because most, if not all, radiations have 
some direct or indirect chemical effect. This diagram is 
supplemented by a verbal summary in Table I. Although 
the magnitude of the unit of wavelength, mp, which is used 
here is of little interest, some readers may wish to attempt 
to visualize its very small value or to refresh their memory. 
It is one millionth of a millimeter. 
TABLE I 
Some Facts Pertaining to Radiation 
Wavelength in m/x 
to 200 
Extreme ultraviolet region 
Absorbed by most media 
200 
to 300 
Middle ultraviolet region 
290 
310 
300 
to 400 
Transmitted by quartz and water 
Absorbed by ordinary glasses and ozone 
Destroys animal tissue 
Destroys germs 
Approximate short-wave limit of solar radiation at the 
earth, the radiation of shorter wavelength probably 
being absorbed by ozone in upper atmosphere. 
Approximate long-wave limit of germicidal action 
Near ultraviolet region 
390 
to 760 
Transmitted by quartz, water and ordinary glasses 
Visible radiation or radiations causing sensation of light 
390 
to 430 
and color 
Violet region 
430 
to 470 
Blue region 
470 
to 500 
Blue-green region 
500 
to 530 
Green region 
530 
to 560 
Yellow-green region 
(See also Fig. 1) 14 
LIGHT AND WORK 
Wavelength in m/j. 
560 to 590 
Yellow region 
590 to 620 
Orange region 
620 to 760 
Red region 
460 
Maximum energy in blue-sky radiation 
555 
Maximum sensibility of the eye 
555 (approx) Maximum energy in solar radiation 
760 to ... 
Infra-red radiation 
Absorbed by thick layers of water 
Partially absorbed by water-vapor in atmosphere 
Biological Effects of Radiation 
Let us return for the moment to daylight in order not 
to lose sight of the conditions under which civilized man 
evolved. Complete ignorance could not exist where ex- 
perience was remembered. This was the situation with early 
primitive beings. Gradually observation developed and 
then finally experimentation which has done so much in 
harnessing natural phenomena for our specific uses.2 The 
present-day savage is not without some knowledge of the 
properties of solar radiation. In a vague way he knows 
that it helps to heal wounds and to make water safer to 
drink. He knows nothing of germs or of therapeutics. Per- 
haps instinct and vague knowledge cause him to utilize to 
some extent the beneficial effects of solar radiation, but does 
he suspect the consequence of the absence of sunlight par- 
ticularly for one who is unfortified by an inclination to 
substitute cleanliness for the protection of solar radiation? 
Modern man has learned many facts of radiation, but much 
remains to be unravelled in the realm of radiation and light 
therapeutics. 
Sunbaths are known to be efficacious in certain cases. As 
a consequence mankind has developed baths of artificial 
radiation. Sometimes only the light is considered and this 
can be of various colors but there is always the invisible 
radiation which may be playing a part. The germicidal 
activity of solar radiation is known to be due to the ultra- 
violet radiation. Rain falls over the earth and it runs off 
to streams, rivers and lakes, carrying with it vegetable mat- LIGHT AND LIFE 
15 
ter and the germs of animal disease and waste. In sunlight 
these germs are killed so that we have outdoors an exten- 
sive sterilizing process operating on land and in the water. 
Hence, we have our open reservoirs where the sun's aid is 
invited in sterilizing the water for our towns and cities. 
Too much solar radiation is harmful but on the other 
hand many diseases show a seasonal variation with maxi- 
mum prevalence in months of diminished sunlight. For ex- 
ample, rickets are more prevalent in the winter months than 
in the summer and it is now fairly well proved that the 
low intensity of sunlight and the consequent small amount 
of ultraviolet radiation is responsible for the increase in 
rickets in winter. In many other ways solar radiation is a 
protector and a servant. In this age of artificial light we 
must consider the composition and intensity of daylight. 
Without other knowledge we may safely assume that arti- 
ficial light which approximates the characteristics of day- 
light is best. The radiations found in daylight should be 
harmless when they accompany artificial light. Natural 
lighting conditions should be best suited to life and there- 
fore should be approached if possible by artificial lighting. 
Of course, adequate knowledge may lead to improvement 
over Nature in specific requirements arising from the arti- 
ficial conditions which mankind has produced. 
Fundamental Lighting Conditions 
If artificial lighting conditions differ radically from 
natural lighting conditions in such matters as intensity, 
spectral character, distribution, and accompanying invisible 
radiations, they must be questioned until adequate knowl- 
edge of the psycho-physiological is available. Human 
beings are still Nature's animals in respect to the great 
environmental factors. From the viewpoint of evolution we 
have just come indoors. We must still view our welfare in 
terms of that environment in which we evolved. Of course, 16 
LIGHT AND WORK 
natural conditions are a resultant or a compromise well 
suited to the complexity of requirements of outdoor life. It 
seems reasonable to suppose that man can improve upon 
these conditions in specific cases. Perhaps he may finally 
acquire the knowledge necessary to improve upon natural 
conditions in every specific respect. However, many 
environmental factors which require countless generations 
to show their effects, must be reckoned with. We cannot 
expect our visual organs and physiological processes to re- 
concile themselves in a day or a year or a century to radical 
changes in a foundational factor such as lighting. 
Fortunately, the human being is fortified with a certain 
resistance to harmful conditions. The human body is con- 
stantly repairing damages. The mental process of man is 
adaptive and can ignore unnatural conditions to some ex- 
tent. But is not the best environment that which would 
permit most of the energy required for repairs to be utilized 
in material and mental output? A factory, for example, 
in which friction and other forms of waste are reduced, will 
produce more at the same expenditure of horse-power and 
man-power than the wasteful factory. Lighting which is 
inadequate or improper, or both, produces friction in many 
ways. It decreases the output whether the factory or office 
manager knows it or not. Furthermore, a light-source just 
here and there does not produce a condition conducive to 
the best production. It is too commonly believed that if 
darkness is dispelled the lighting problem is solved. Let us 
view the span from darkness to the natural lighting condi- 
tions outdoors on a reasonably clear day. The one extreme 
represents a condition of little or no activity possible. The 
other extreme is that under which we as active beings 
evolved. Sometimes the intensity is too great for serious 
visual work, such as reading, but this intensity is not likely 
to be approached in general artificial lighting. However, 
viewed in this manner we find ordinary intensities of arti- LIGHT AND LIFE 
ficial illumination hovering near the darkness extreme. 
From every viewpoint, available data show that great ad- 
vantages would accrue if artificial lighting conditions were 
further away from the darkness extreme and closer to 
natural lighting conditions outdoors. 
17 
What Good Lighting Accomplishes 
Further discussions of these factors will be found in chap- 
ters devoted to the specific phases. In closing this intro- 
duction let us consider what adequate and proper lighting 
may accomplish in human activities. 
a. Certainly the production of a factory in darkness 
would be small and uncertain. It increases with 
increasing light. Why should we expect it to 
reach a maximum under conditions now preva- 
lent when these can be improved in so many 
ways. Experiments point conclusively to greater 
production with better lighting and the value 
of this production is many times greater than the 
additional cost of lighting. 
b. We cannot expect good workmanship in the dark. 
The best workmanship certainly is not achieved 
even under poor lighting. Evidence points to 
increasing quality of work as we improve the 
lighting. Particularly for fine work, very much 
higher intensities than are now prevalent will 
pay dividends. The lighting must be proper as 
well as adequate. 
c. We would expect workmen to spoil much material 
in the dark. This spoilage is reduced even with 
poor lighting but it is not a minimum until the 
lighting is adequate and proper. 
d. Safety is certainly not a product of darkness. Ade- 
quate and proper lighting increase the safety 
of the workmen. 18 
LIGHT AND WORK 
e. The eyes are rested in the dark and therefore eye- 
strain and eye-fatigue are not produced in dark- 
ness. They are the consequence of inadequate 
and improper lighting. Certainly our eyes 
which have evolved under natural lighting con- 
ditions, would be less strained and fatigued 
under artificial lighting conditions approxima- 
ting these in quality, quantity, and distribution 
of light and brightness. 
j. Darkness is the period of rest. Light is stimulat- 
ing. Adequate and proper lighting stimulate 
greater activity which, properly directed, results 
in greater production. 
g. Well-lighted surroundings promote cheerfulness 
and workmen are not likely to become too cheer- 
ful from light alone. Therefore, there is no dan- 
ger of overlighting in this respect. Certainly 
workmen are depressed by improper and inade- 
quate lighting. 
h. We would not expect orderliness and good dis- 
cipline to be best in darkness and we have no 
reason to expect them to reach their highest 
development excepting under adequate and 
proper lighting. 
Doubtless there are many minor factors which could be 
mentioned, but these chief ones are sufficient to indicate why 
good lighting is an asset in factories, offices, stores, and 
homes; in fact, wherever persons work or use their eyes in 
serious work or wherever there is traffic. Therefore, light 
is not only essential to human beings who depend so much 
upon vision, but is a great economic factor. The best will 
not be obtained from light until the lighting conditions are 
adequate and proper and that means they must approach 
natural lighting in certain fundamental aspects. LIGHT AND LIFE 
19 
Unfortunately light is not a novelty; in fact, it is a com- 
mon experience from the time we are born and throughout 
our lives. For this reason it is accepted as a matter of 
course and no serious thought is given to it by most persons. 
It is almost as omnipresent as the air we breathe. But we 
are learning that plenty of fresh air is healthful; perhaps 
we will learn that good lighting is just as import- 
ant to us who depend so much upon vision. Most of our 
experiences and activities depend upon the doorway of 
vision and therefore light and lighting are intricately asso- 
ciated with most of our accomplishments and other ex- 
periences. 
Most persons whose consciousness has been awakened in 
some manner to light and lighting, doubtless will not sense 
the important difference between bad and good lighting. 
It may not be difficult for them to realize the differences in 
direct results. They may readily admit that glaring lights 
or feeble illumination may increase the likelihood of acci- 
dents, but it is not so easy to sense the physiological and psy- 
chological effects. These are less direct in their manifesta- 
tions. The entire psycho-physiological realm is not as sus- 
ceptible to analysis and appraisal. The eye endures much 
abuse without complaining directly, but all these abuses of 
inadequate and improper lighting are recorded eventually 
on the loss side of the human ledger. We cannot, without 
risk, subject our sense-organs and mental being to condi- 
tions radically different from those for which they have 
become tuned. This is one of the premises of this book 
and it is the aim of chapters which follow to discuss the 
lighting conditions of Nature and to indicate how they can 
be approximated and even improved upon by means of 
artificial light. Chapter II 
DAYLIGHT OUTDOORS 
TO civilized people daylight outdoors is apparently of 
chief interest from the viewpoint of its influence upon 
daylighting indoors. However, from the point of view of 
the relation of lighting to human activity - production, 
safety, and happiness - natural lighting conditions out- 
doors are of great fundamental importance. As has been 
pointed out, these were a prominent part of the environment 
to which our body, sense-organs, and mind became attuned 
through ages of evolution. On coming indoors we cannot 
wipe out these imprints in a few generations, although some 
of them may not be as deep or as resistant as others. Doubt- 
less some of them are now being altered by the changes 
which civilization has introduced and they may be obliter- 
ated in time. However, from what is known of evolution 
and environment, it is safe to state that most of the im- 
prints of outdoor environment are still plainly visible and 
that they will remain so for a long time. 
The chief factors in the distribution of light and of 
brightness outdoors are the sun, the sky, the atmosphere, 
the clouds, and the colors and the reflection-factors of va- 
rious areas. These are important individually and in com- 
binations and their influences are far-reaching. 
Sunlight 
Notwithstanding the sun's great distance, the intensity 
of illumination at the earth's surface due to it is enormous. 
That most persons are unconscious of the difference be- 
tween the intensity of sunlight illumination and that ordi- 
narily encountered in artificial lighting, is evidence of the 
20 DAYLIGHT OUTDOORS 
21 
extreme range of sensibility of our visual sense-organ and 
its attendant psycho-physiological processes. The intensity 
of illumination due to direct sunlight on a very clear day 
when the sun is near the zenith is nearly 10,000 foot- 
candles. Inasmuch as a foot-candle is the illumination pro- 
duced by a standard candle (about one inch in diameter) 
at a distance of one foot, it is seen that we must have the 
equivalent of 10,000 burning candles concentrated at a point 
one foot from a surface in order to equal this intensity of 
sunlight illumination. Only one-tenth of this intensity of 
illumination is obtained on a surface by means of a 1,000- 
watt gas-filled tungsten lamp even when the surface is 
placed as close as one foot to the filament. Taking into 
account the distance of 93 million miles to the sun and the 
fact that the illumination varies inversely as the square of the 
distance, we find that the sun is a light-source equivalent to 
2.5 billion billion billion candles. This is only of passing 
interest; however, it emphasizes not only the tremendous 
scale upon which the universe is built, but also the great 
magnitude of natural lighting compared with which man's 
achievements so far are feeble indeed. 
The candle-power of a light-source in a certain direction 
is due to its brightness and to the area of this brightness 
in the particular direction. The sun is not only of great 
area, but it is much brighter than any light-source that man 
has devised for lighting purposes. It is true that by " ex- 
ploding " wires by means of electric current far in excess of 
their safe carrying-capacity, temperatures equalling, and 
even far exceeding, that of the sun have been produced. 
But among the present-day practicable light-sources there 
is none which approaches the brightness of the sun although 
this is one of the goals of light-production. Inasmuch as 
the spectral character of light is closely associated with 
brightness and temperature of a hot body, other means have 
been adopted by man to reproduce sunlight and skylight 22 
LIGHT AND WORK 
artificially. These various phases are discussed in later 
chapters. 
There is a natural variation in the intensity of sunlight 
illumination at any point on the earth due to the variation 
in the altitude of the sun. For example, if a white surface 
is held perpendicular to a " stream " of light-rays from a 
source it will be seen to have a certain brightness. If it is 
now tilted somewhat the brightness will diminish. In other 
words, the illumination intensity depends upon the density 
10000 
INTENSITY IN TOOT-CANDLES 
8000 
6000 
4000 
zooc 
4 5 6 7 8 3 10 II 12 I Z345678 
HOUR 
Fig. 2. Variation of skylight, sunlight and total daylight on a horizontal 
surface throughout a typical clear day in midsummer. 
of light per unit of area of the surface. A horizontal sur- 
face receives no direct sunlight at the instant of sunrise 
and it receives a maximum amount of direct sunlight at 
noon on a clear day assuming all other conditions are 
constant. 
The variations in the quantity of light3 and of total ra- 
diation 4 emitted by the sun have been extensively studied. 
At most they are only a few per cent and therefore are of 
no importance from a lighting viewpont, although they are 
of considerable scientific interest in many ways. DAYLIGHT OUTDOORS 
23 
From a lighting viewpoint we are chiefly interested in the 
quality and the great intensity of sunlight, in the variation 
in the sunlight illumination due to latitude, sky-conditions, 
the altitude of the sun and in the single shadow which it 
casts. 
The general relations of intensity of illumination on sur- 
faces horizontal, or nearly so, due to sunlight and to sky- 
light throughout a typical clear day in midsummer, are 
shown in Fig. 2 for the middle latitudes. If the total illu- 
mination intensity at noon is 10,000 foot-candles, it is seen 
that artificial lighting intensities, which are commonly only 
24 
H0UR5 DURATION OF 5UN5HINE 
20 
16 
IZ 
8 
4 
0 
10 
20 
30 
40 
50 
60 
70 
LATITUDE IN DEGREES 
Fig. 3. The maximum possible duration of sunlight in midsummer 
and in midwinter for different latitudes in the northern hemisphere. 
a few foot-candles, would be represented by only a portion 
of the thickness of the base line. Twilights are represented 
by the skylight which appears before sunrise and persists 
after sunset. On a typical clear day in winter the relations 
are similar but the intensities of illumination are less owing 
to the generally lower altitude of the sun. 24 
LIGHT AND WORK 
The maximum duration of direct sunlight on a horizontal 
surface (interval between sunrise and sunset) varies, of 
course, with the season and the latitude. These are shown 
for midsummer and midwinter in Fig. 3. 
Direct sunlight at the earth's surface on a clear day in 
middle latitudes, when the sun is at its maximum altitude, 
varies generally from about 7,000 foot-candles in winter 
to about 9,000 foot-candles in summer on a surface per- 
pendicular to the sun's rays. Owing to the inclination of 
the sun's rays, the intensities on a horizontal surface are 
slightly less than those for normal incidence presented 
above. Throughout the entire year on clear days the total 
daylight illumination intensities on surfaces not more than 
thirty degrees from horizontal are generally greater than 
1,000 foot-candles excepting during the hour after sunrise 
and that before sunset. Throughout most of the day on 
fairly clear days, the total daylight intensity of illumination 
is several thousand foot-candles, approaching 10,000 foot- 
candles for two hours during midday. Of this total day- 
light on clear days, that due to direct sunlight is about 80 
per cent of the total. Even on surfaces nearly vertical and 
facing north (exposure to only a portion of the sky and not 
to the direct sun) the intensity of illumination, during most 
of the day throughout the year, is generally several hundred 
foot-candles. 
Effects of the Atmosphere 
From the viewpoint of light and of radiation, the atmos- 
phere performs important functions. We live at the bottom 
of this ocean of atmosphere and are adapted to its presence 
just as fish in their " oceans " of water are adapted to the 
properties and influences of water. Even though we sup- 
plied ourselves with oxygen from another source and could 
likewise supply plant-life with the constituents of the atmos- 
phere which it needed, all life as we know it would quickly DAYLIGHT OUTDOORS 
25 
disappear if the atmosphere ceased to exist. No better 
example of the adaptation of life to environment is avail- 
able. The protection of the atmosphere is necessary be- 
cause we are used to it. From the viewpoint of lighting we 
are interested in the atmosphere chiefly because it filters 
solar radiation, scatters light, and causes great variations 
in daylight intensities. (See Fig. i and Table I) 
In general, any material body or fluid is not transparent 
to radiant energy of all wavelengths. Clear glass is quite 
transparent to visible radiation or light but is opaque to 
ultraviolet radiation of the " middle " ultraviolet region 
(wavelengths shorter than 300 mp). Likewise the gases 
of the atmosphere 5 are transparent to visible radiation but 
they absorb invisible radiations (ultraviolet and infra-red) 
of certain wavelengths. Water-vapor, carbon dioxide, and 
ozone absorb infra-red radiation, although most of this ab- 
sorption is attributed to water-vapor in the layer of atmos- 
phere near the earth. Infra-red radiation is of no use to 
us from the viewpoint of lighting and vision and, therefore, 
that which is absorbed by constituents of the atmosphere 
is of no direct interest. However, infra-red radiation is 
absorbed by the eye-media and too much of it may be harm- 
ful, although there is no experimental evidence of any 
harmful effect of infra-red radiation under ordinary natural 
or artificial lighting conditions. 
It is worth noting that the absorption of the atmosphere 
does reduce the amount of radiant energy accompanying 
the daylight which enters the eye. Inasmuch as present- 
day artificial light-sources emit relatively much greater 
quantities of infra-red radiation per unit of light than the 
sun (or the sky), much more heating effect results from 
equal illumination intensities of artificial light than from 
natural light. This presents a formidable difficulty in simu- 
lating natural lighting conditions by means of present-day 
artificial light-sources. For example, some provision must 26 
LIGHT AND WORK 
be made, such as a layer of water for screening out the infra- 
red radiation, when hundreds or thousands of foot-candles 
of artificial illumination are desired as an adequate substi- 
tute for sunlight, as in the case in experimenting in 
plant growth, therapy, dye-fading, etc. However, even with 
present-day artificial light-sources this problem will not be 
met in ordinary lighting problems until the prevalent in- 
tensities are increased many fold. Therefore, great im- 
provements can now be made without this difficulty 
and it is likely that eventually " colder " light will be 
produced. This is the goal of the scientist in light- 
production. 
In the ultraviolet region the absorption of the atmosphere 
is very important. A body as hot as the sun emits ultra- 
violet radiations very generously, throughout the near, 
middle, and extreme regions. However, for many years it 
was a mystery why solar radiation reaching the earth did 
not contain energy of wavelengths appreciably shorter than 
290 mp in the ultraviolet region. The human body is not 
adapted to these radiations which are absent; hence, it 
is not surprising to find them irritable to and even destruc- 
tive of pigment cells, eye-media, etc. Various artificial 
light-sources, such as the carbon arc, the magnetite arc, and 
the quartz mercury arc, emit ultraviolet radiations shorter 
in wavelength than the shortest found in daylight and these 
are harmful to the eyes. Many serious results have been 
due to carelessness and to ignorance, but when these sources 
are surrounded by ordinary glass, harmful amounts of ultra- 
violet radiation are not transmitted. Even with the protec- 
tion of glass one should not look steadily at the light-source. 
The mystery of the absence of the ultraviolet radiation of 
the shorter wavelengths from solar radiation which reaches 
the lower regions of the atmosphere has been explained 
by attributing the absorption of them to the ozone in the 
extreme upper regions of the atmosphere. DAYLIGHT OUTDOORS 
27 
These ultraviolet radiations to which we are not adapted 
and which, therefore, are harmful have good uses. Here man 
extends his independence from natural conditions by manu- 
facturing what is not found in Nature. He uses these radia- 
tions, which are harmful to the eyes, in killing germs, in 
therapeutics and in many chemical processes. Fortunately, 
germs succumb to ultraviolet radiation as long as 310 mp, 
in wavelength and this narrow range of germicidal rays 
found in solar radiation is efficacious owing to the great 
abundance of daylight. The same is true of other biological 
effects of radiation. Our skin is tanned but quickly develops 
or loses resistance because of the superficiality of the effect 
in the pigment cells. 
It is seen that the atmosphere by absorption shortens the 
extent of the spectrum of solar radiation in the ultraviolet 
region and reduces materially the amount of infra-red ra- 
diation without appreciable absorption of visible radiation. 
Thus we have a human eye most sensitive to the range of 
wavelengths of the most plentiful radiation. No other 
result of adaptation to environment could be expected. 
There are many other important parts that the atmos- 
phere plays in respect to life on earth, but the only other 
one of primary interest here is the scattering of light by it. 
This is what produces skylight. 
Clouds, of course, produce the greatest variation in the 
intensity of daylight illumination at any given time and 
place. These variations are enormous; in fact, at any hour 
of the day or year they may cover a range from zero day- 
light to the maximum possible. This great variation is the 
most undesirable feature of daylight from the viewpoint of 
our activities indoors. The variations of direct sunlight 
on a horizontal surface outdoors are indicated for a few 
typical days 6 in Fig. 4. In this respect man is able to im- 
prove upon daylight by means of artificial light. However, 
the great variation is less noticeable and less annoying owing 28 
LIGHT AND WORK 
to the great intensities of daylight than it would be in the 
case of ordinary artificial lighting intensities. Such varia- 
tions would be dangerous, very annoying and even unbear- 
able in the case of prevalent artificial-lighting levels of 
illumination. 
Skylight 
Apparently every medium scatters light, even the most 
transparent solid, liquid, or gas. The purest water of great 
depth appears bluish green and it has some brightness only 
ID 
This top line represents "no direct sunlight" 
£loudy sky 
slightly broken 
RATIO or 5KYLIGHT TO TOTAL DAYLIGHT 
.9 
3 
7 
.6 
.5 
4 
.3 
.2 
.1 
-Thinly over- 
cast sky 
i \ 
zThin 
clouds 
• Dense 
clouds 
-Clear sky 
some clouds 
• Clear sky 
This base line represents "no skylight' 
9 10 11 1Z 1 2 3 4 5 
TIME OF DAY 
Fig. 4. Showing the variations in the ratio of the intensity of skylight 
illumination to that of total daylight illumination on a horizontal surface 
throughout four typical days. These different days may be described respec- 
tively as dense, slightly broken clouds; thinly overcast sky; clear day with 
few clouds; and clear day. 
accounted for by scattered light. This is true of the atmos- 
phere, although, as seen from the earth's surface, most of 
the sky's brightness on a clear day is due to the illuminated 
particles of dust, smoke, and water. Of course, atmos- 
pheric conditions are changing continually but, for example, 
measurements made by the author on a certain clear morn- 
ing showed that the sky was ten times brighter as seen from DAYLIGHT OUTDOORS 
29 
the earth than from altitudes of ten to twenty thousand feet. 
The scattering in the lower layers is due largely to solid or 
liquid " foreign " particles, but that of the high regions is 
perhaps chiefly due to molecules of gases. Certainly little 
or no water-vapor (in the form of ice crystals) can remain 
suspended in the extreme upper regions even if it should get 
there occasionally. 
As one proceeds upward the sky gets darker and darker, 
so, to one who has had this experience, it is obvious that the 
sky is bright merely because it scatters sunlight. In fact, 
if water, smoke, dust, and ice particles were absent and the 
atmosphere was purely gaseous, the sky would be a very 
dark blue. If the atmosphere did not exist at all and the 
void of outer space began at the earth's surface, we would 
have a brilliantly harsh sun shining in a " night " sky 
studded with stars. Adapted as we are to a bright sky, its 
absence would leave a most unbearable lighting effect which 
begins to be suggested at very high altitudes (when there 
are no cirrus clouds) and is partially demonstrated by a 
street-light or lantern outdoors at night at a considerable 
distance from objects such as trees and houses. Cover the 
ceiling and walls of a room with black velvet and one has 
a fair reproduction of the condition which would be un- 
bearable to us adapted as we are to the bright sky. 
The blue color of the sky is due to the fact that very 
small particles and molecules scatter the light of short wave- 
lengths (and particularly blue light) more than light of 
longer wavelengths. The direct sunlight, being robbed of 
more of these violet, blue, and green rays than of the yel- 
low, orange, and red rays, appears more reddish than it 
would at the outer layers of atmosphere. As the sun de- 
creases in altitude, it penetrates greater and greater layers 
of atmosphere, hence the orange or red setting sun. The 
variation in the color of the setting sun is due largely to the 
variation in the amounts of smoke, dust, water, and ice 30 
LIGHT AND WORK 
particles and total depth of atmosphere. (See Figs. 18 
and 19) 
An interesting demonstration of the selective scattering 
of light by small particles may be made with a lighted cigar. 
The smoke curling upward from the end is decidedly bluish. 
That expelled from the mouth is more nearly white. The 
particles in the latter case are larger owing to the condensa- 
tion of moisture upon them. The larger particles do not 
scatter the light of short wavelengths as predominantly as 
the smaller particles do. The shadow cast by the cloud of 
smoke on a white paper is of a ruddy color showing that it 
is selective in transmission at least, thus providing a dem- 
onstration of the chief reason for the red appearance of 
the setting sun. 
These various factors which give us a blue sky and a red- 
dish sun at low altitudes are important. We are particularly 
interested in the color or quality of daylight which as we 
have seen is variable and due to a combination of lights 
from the sun, the sky, clouds, etc. Quality of light is dis- 
cussed in Chapters V and VI. 
brom a lighting viewpoint we are even more concerned 
with the lighting effects due to the large area of sky. This 
not only subdues the harshness of the sun but also of the 
lighting effect. The sky sends light into the shadows and 
by brightening them renders them less harsh. On an aver- 
age clear day the brightness of a shadow cast by an object 
in direct sunlight during midday, is about one-fifth as bright 
as the sunlit surface when the shadow is receiving light from 
an entire sky. As the sky " thickens " or grows more hazy, 
the brightness of the shadow increases until it no longer 
exists. The approximate percentages of skylight and sun- 
light for various sky conditions are shown in Table II. DAYLIGHT OUTDOORS 
31 
TABLE II 
Percentages of direct sunlight and of skylight in the 
total light reaching a horizontal surface at noon for 
various sky-conditions. 
Skylight 
Sunlight 
Outside of atmosphere 
o 
100 
Very clear day 
IO 
90 
Average clear day 
20 
80 
Very thin haze 
3° 
7° 
Moderate haze 
5° 
50 
Thickly overcast 
100 
0 
The ratios of total skylight illumination on a horizontal 
plane at the earth's surface to the illumination due to 
total daylight are shown in Fig. 4 for four typical summer 
days in a middle latitude. These show that the illumination 
due to skylight is less variable than that due to direct sun- 
light. Although the average illumination intensity due to 
skylight varies somewhat, the illustration serves very well 
to indicate the extreme variation in direct sunlight due to 
moving clouds and to haziness of the sky. 
On partially cloudy days the percentage of light from the 
sky is extremely variable but it averages somewhere near 
the conditions obtaining when there is a thin haze. 
As haziness or cloudiness increases, skylight becomes an 
increasing percentage of the total until when the sun is 
obscured, obviously all the light comes from the sky. As 
the sky brightens, the total light reaching the earth dimin- 
ishes so that outdoors the maximum intensities of illumina- 
tion are found on clear days. However, this is not true 
indoors where the sky is more important as a light-source 
than the sun. Hence, indoors the highest intensities of day- 
light, where skylight is depended upon solely, as it usually 
is, are found on days when the sky is the brightest. The 
moderately hazy skies are the brightest and, therefore, are 
the best from the viewpoint of indoor daylighting. 
The actual brightness of the sky is of importance both in 
computing the intensities of illumination obtainable from 
given sky-areas and fundamentally in obtaining an idea of 32 
LIGHT AND WORK 
the brightnesses to which we have become adapted. Bright- 
ness can be expressed in several ways, but one of the 
simplest is in terms of candles per square inch. If a square- 
inch opening is cut in a cardboard and the sky is viewed 
through it, the luminous intensity or candle-power of this 
light-source (the square-inch of sky-brightness) can be de- 
termined as easily as that of any light-source. The sky- 
brightness varies considerably but it is usually less than four 
candles per square inch. An attempt has been made to ex- 
press the approximate average brightnesses of the sky 
under various conditions in Table III. 
TABLE III 
Average brightness of the sky under various conditions. 
Candles per 
square inch 
Outside of atmosphere 
... 0.0 
Very clear day 
... 0.5 
Ordinary blue sky 
... 1.0 
Thin haze 
... 2.0 
Moderate haze 
... 30 
Dense haze or thin cloud layer 
... 4.0 
Medium cloud layer 
... 2.0 
Thickly overcast 
... 0.5 
Densest overcast approaches 
... 0.0 
Maximum brightness of sunlit clouds 
... 20.0 
The sunlit surface of clouds varies in brightness, de- 
pending upon the thickness of the clouds. Thick clouds 
have been found by the author 8 to reflect as much as 78 
per cent of the incident light. It is readily computed that 
the maximum brightness of sunlit clouds is approximately 
20 candles per square inch. The sunlit portions of clouds 
are usually from 5 to 10 times brighter than an adjacent 
area of sky. 
These brightnesses of the sky and of the clouds are inter- 
esting from the viewpoint of their relations to illumination 
intensities outdoors and indoors. They also indicate the 
approximate maximum brightness to which the human eye 
was commonly subjected throughout the ages. Under the 
lighting conditions outdoors, the bright areas of clouds and DAYLIGHT OUTDOORS 
33 
sky can be viewed with comfort and they appear to be 
safe maximum brightnesses in the consideration of artificial 
lighting. Of course, these brightnesses indoors can be very 
annoying if the lighting conditions (level of illumination, 
adjacent brightnesses, etc.) are greatly different from those 
outdoors as they often are. 
Without skylight we would have no periods of twilight 
before sunrise and after sunset. Day would come abruptly 
and night would close down almost instantly. Accustomed 
to these modulations from night to day and from day to 
night, skylight is playing an important role in natural light- 
ing. In the foregoing we find fundamental lessons in light- 
ing which must be taken into account in developing a science 
of artificial lighting. Natural lighting conditions teach us 
that large areas of relatively low brightness are desirable 
adjuncts to direct light in reducing the harshness of direct 
lighting and in providing light which penetrates shadows. 
Twilight periods also remind us that our eyes are used to 
modulations in intensities of light, that is, that our visual 
organs have been accustomed to certain time-intervals in 
which to become adapted to extreme changes. 
Daylight intensities are usually less than 100 foot-candles 
at sunrise and sunset, but they are much greater than maxi- 
mum artificial illumination intensities in general use for 
some time before sunrise and after sunset. When twilight 
has decreased to a few foot-candles we usually think it is 
too dark for serious visual work. After turning on the arti- 
ficial light we generally experience annoyance or a feeling 
of inadequate illumination for some time. This feeling per- 
sists until daylight is practically zero and our eyes are 
adapted to the low level of artificial illumination. Usually 
twilight becomes insufficient for outdoor activities when the 
sun is six degrees below the horizon. This is about a half- 
hour before sunrise and after sunset in middle latitudes.7 
Smoke has a very marked effect on the intensity of day- LIGHT AND WORK 
34 
light illumination. In smoky cities, direct sunlight and 
total daylight are often decreased 50 per cent by smoke. 
If not too dense, smoke causes the sky to be brighter than 
when the atmosphere is clear but, of course, even then the 
net result is a marked decrease in the total amount of day- 
light reaching the earth. 
The amount of scattered sunlight is greatest near the 
location of the sun in the sky. For this reason the sky- 
16 
14 
Jojt 
12 
io 
RELATIVE BRIGHTNESS OF SKY 
e 
■west 
6 
4 
South 
"Rorth 
2 
0 
9 
IO 
II 
A. M 
12 
TIME OF DAY 
I 
2 
3 
P.M. 
4 
5 
Fig- 5- Variations in the brightness of patches of sky about 20 de- 
grees above the horizon in the north, east, south, and west respectively, 
throughout most of a clear day in August at 40 deg. north latitude. 
brightness is much greater in this region and the sky as a 
source of light is much more variable in those regions 
traversed by the sun's path than in other parts. This is 
well illustrated in Fig. 5 where the variations in the bright- 
ness of a patch of sky about 20 degrees above the horizon 
at four points of the compass - north, east, south, and west DAYLIGHT OUTDOORS 
35 
- are shown for a clear day. The measurements were made 
in a middle north latitude (40 deg. N) between 9 a.m. and 
5 p.m. It is seen that the eastern sky was brighter in the 
morning than in the afternoon after the sun had moved 
into the west. Conversely the western patch of sky was 
brightest in the afternoon. Thus the eastern and western 
skies are seen to vary greatly in brightness. The north and 
south skies are much less variable in brightness (exclusive 
of clouds) and of the four the north sky is more nearly con- 
stant in brightness than the others. 
This reminds us that the north sky has been chosen by 
colorists, painters, photographers and others who had need 
for daylight as constant as possible. In developing artificial 
lighting we should remember that north skylight has been 
used for centuries largely for its steadiness rather than for 
its spectral character or color. There are many obvious 
cases where light of a quality of noon sunlight is better 
than that of north skylight, but it is difficult to convince 
those who have used north skylight as a standard that this 
is true. Having been driven to a north window because of a 
need for reasonably steady light, they have become used 
to the quality or spectral character of north skylight and 
have forgotten or never knew that it was not the spectral 
character but the constancy of north skylight that made 
them turn to the north. Incidentally the north skylight also 
varies less in spectral character than any other areas of 
sky in a north latitude. This is further discussed in 
Chapter V. 
Moonlight and Starlight 
If there were no moon or stars it is likely that there would 
not have remained in the human eye the receptors which 
are sensitive only to very feeble light. After twilight failed 
completely we would have only the suspicion of light ex- 
tremely faint in comparison with starlight which is itself 36 
LIGHT AND WORK 
so weak as to be almost inexpressible in comparison with 
daylight. Nature has a way of eliminating useless things 
and it is likely that our retinal receptors - termed " rods " 
because of their shape - which enable us to see in moon- 
light and starlight, would have vanished. We say vanished 
because it is possible that they might have been inherited 
from animal ancestry which needed them. Prehistoric or 
very early animal life lived in the sea where amid its dark 
depths it was advantageous to have receptors which were 
sensitive to feeble light. Other animals may have needed 
these sensitive receptors owing to their life in caves. So we 
cannot say just what the human eye inherited from the past. 
At any rate moonlight and starlight are a part of Nature 
and we can see under these faint illuminations. Moonlight 
intensities of illumination are of the order of magnitude 
of a few hundredths of a foot-candle, a matter of a few mil- 
lionths of maximum intensities of daylight. Starlight is of 
the order of magnitude of a billionth of maximum daylight. 
Of course, our ability to see under these natural night 
illuminants is much more restricted than under the high in- 
tensities of daylight. We cannot distinguish color at all and 
we cannot see fine detail or brightness differences that we 
can see in the daytime. Nevertheless, the fact that we can 
see at all under illumination intensities only a millionth or 
a billionth of maximum daylight marks the human visual 
process as a marvel in range of sensibility. 
These very low intensities were of use to primitive man 
and are also to us but they are not safe intensities for the 
complex activities of civilization. We cannot see with suf- 
ficient certainty to engage the eyes in serious work. In fact, 
the chief interest is that we can see at all. However, moon- 
light is a good example of direct lighting without any ap- 
preciable apparent general lighting from the sky. It is like 
sunlight without much skylight, for the brightness of the sky 
at full moon is so feeble as to contribute scarcely any appar- DAYLIGHT OUTDOORS 
37 
ent light in the shadows. What light we see in the shadows 
is chiefly reflected from nearby objects. As a consequence 
we have harsh lighting effects. 
The color of moonlight is practically the same as noon 
sunlight notwithstanding many opinions to the contrary. 
On the stage moonlight is commonly represented as of a 
conspicuous blue-green hue. In reality it is practically 
white. It is sunlight reflected from the moon's surface which 
is practically colorless. 
Nature's Landscapes 
The distribution of brightnesses and of colors in Nature 
is important from a lighting viewpoint. In every respect the 
human being became adapted to these and one would find it 
difficult to defend any radically different conditions as being 
better to live with. The predominant light-source, the sun, 
is so very bright that we cannot look at it without serious 
discomfort. Hence, we avoid it and Nature has helped us 
to avoid it readily by providing us with " visors." We go 
further in this respect and provide extensions on our hats 
and caps to screen the eyes from sun and sky. The latter 
is of a brightness which does not leave serious blinding after- 
images and it may be looked at directly outdoors without 
discomfort. The general high level of brightnesses makes the 
brightness of the sky endurable outdoors. But indoors 
where the general level of brightnesses is much lower, we 
cannot view contrasts and brightnesses of the same magni- 
tude as outdoors with the same comfort. Still, indoors we 
find little attention paid to these matters and mankind is 
responsible for bare light-sources of very high brightness, 
extreme brightness-contrasts and low levels of illumination, 
all contrary to Nature and, therefore, contrary to conditions 
to which our psycho-physiological processes are adapted. 
The fact that we have very generally adopted decorative 
schemes to live with, in which the brightnesses increase from LIGHT AND WORK 
38 
floor to ceiling, is significant. This is like Nature. Some 
may say that such a distribution of brightnesses is logical 
because it is more stable. That is, the dark floor seems to 
support the brighter walls and these in turn seem to support 
the brighter ceiling. But, we may ask, why does such a 
decorative scheme appear more stable? Why does the 
darker seem to support the brighter? Surely we can find 
the answer only in natural surroundings which not only made 
us physically what we are but also what we are physio- 
logically and psychologically. No matter which way we 
turn or how we struggle for independence, we find Nature's 
influences and even shackles. But they are kindly shackles 
if we do not resist them. We must recognize them and be 
guided accordingly. If we desire independence we must 
recognize the fact that we cannot materially hasten Nature's 
processes. They are altered slowly, imperceptibly slowly. 
All the thousands and thousands of years of mankind's his- 
tory since the early primitive men are but a second or two 
in the hour of animal history as written in the geology of 
the earth. 
The reflection-factors or the amounts of light reflected 
by earth areas8 are very small so that in our landscapes 
the foreground and middle distance are of low brightnesses 
compared with the sky. The average range of the amounts 
of perpendicularly incident light reflected by earth areas is 
given in Table IV. 
TABLE IV 
Reflection-Factors of Earth Areas 
Deep clear water 
Shallow inland water . 
5-10 
<c 
u 
Grass and green crops . 
5-10 
u 
U 
Woods 
a c 
u 
Barren soil 
8 2Q 
a 
CC 
Clouds 
cc 
cc 
Snow 
/° 
cc 
cc 
It is difficult to believe that the range of brightnesses 
in a landscape is as great as it is. If the artist were to DAYLIGHT OUTDOORS 
39 
represent the relative brightnesses or " values " accurately, 
he would have to use black for grass and woods and white 
for clouds. It is difficult to believe that woods, for example, 
are only as bright as a flat area painted black but this is 
true, for black paint commonly reflects as much as three 
or four per cent of the light falling perpendicularly upon it. 
The very high intensities of illumination make these low 
reflecting areas seem relatively brighter than they are. In- 
doors with our extremely low levels of illumination we should 
reduce the range of values by using materials having much 
greater reflection-factors than these areas outdoors. 
A sunlit snow-covered landscape is an example of very 
trying conditions. Outdoors such bright areas, below the 
horizontal line or plane of vision, are much less usual than 
areas of low brightness. Over much of the area of the earth 
which is appreciably populated, snow-covered landscapes are 
never seen. In most places occupied by man great snow 
areas are at best relatively rare. They are unnatural. The 
eye is not adapted to high brightnesses below the horizontal 
and they are trying to the eyes. Furthermore, sunlit snow 
is as bright as the brightest sunlit cloud and, therefore, from 
five to ten times brighter than the average brightness of a 
clear sky. Snow also reflects ultraviolet radiation, and, 
although solar radiation which reaches the earth contains 
only a narrow range of wavelengths of ultraviolet energy 
which is irritating to the eyes, the great quantity emitted 
upward into the eyes from sunlit snow is sufficient to do 
damage. Snowblindness is the inevitable result of these 
various factors. 
It is a matter of optics that if a surface could reflect 
all the incident light it would be as bright as a hemispherical 
dome if it received light from all parts of this dome.'1 For 
example, if the sky were uniformly overcast, such a surface 
placed horizontally on the earth's surface would be as bright 
as the sky. If the sky was clear and the sun was contribut- 40 
LIGHT AND WORK 
ing four times as much light to this surface as the sky, the 
surface would be five times brighter than the sky. This is 
practically the case with sunlit snow. However, snow which 
reflected only 80 per cent of the incident light, would be 
only 80 per cent as bright as a uniformly bright overcast 
sky. Thus we see that an object under sky illumination 
alone can only approach the sky in brightness as close as its 
reflection-factor approaches unity. The best mirror can be 
only about 95 per cent as bright as a uniformly bright over- 
cast sky; whitest powder only about 98 per cent, etc. It 
is easy to estimate the brightness of a horizontal surface 
receiving both sunlight and light from the entire sky by 
making allowances for the amount of sunlight falling upon 
it besides reasoning as in the foregoing with the skylight 
illumination. 
Another interesting phase of environmental influence is 
the psycho-physiological imprints that Nature's color- 
scheme has made upon us. This subject is more specifically 
interesting in the home and in decoration in general, but 
it does have some bearing on human activities in work- 
places, such as office and factory. It is difficult to believe 
that colors have any innate powers. It is easier to attribute 
our reactions to them to their location, area, novelty, and 
commonplaceness in the natural environment which made 
us largely what we are.3 It is well known that green is 
more or less neutral or soothing. Why should it not be so? 
Nature s vegetation confronted mankind almost everywhere 
throughout ages of adaptation. That green acquired a neu- 
trality is a most natural consequence of this continued 
adaptation. Therefore, what better color could we choose 
for environment for our daily work in office and factory? 
Blue sky has been with mankind for eons, but in a posi- 
tion not quite as conspicuous. What is more serene than a 
blue sky! Perhaps we feel that it is cool up there. Sky- 
light illuminates shadows and penetrates into Nature's soli- DAYLIGHT OUTDOORS 
41 
tudes. Why should not blue be cool, serene, quiescent, or 
even depressing? 
The sun appears yellowish by contrast with the blue sky. 
Yellow is a bright color. Why should it not suggest light and 
cheerfulness? Yellowish-orange and orange-red were asso- 
ciated with fire ever since mankind used fire and certainly 
that has been for countless centuries. It is not surprising 
that these various hues should be " warm " and stimulating. 
These are barest glimpses into a complex but extremely 
interesting subject.9 They reveal or at least indicate the 
power of environment. We should have some color even for 
the surroundings where we work. Nature is colorful and 
we cannot justify drabness even where we work. When in 
doubt as to the color for walls and other surroundings in office 
and factory we can do worse than to heed Nature's color- 
scheme. Of course, where we specifically aim to utilize the 
powers of colors we may go much further than Nature. 
In calling forth the powers of colors we can get the most 
effect by being different from Nature. We can reverse the 
color-scheme, use large areas of pure colors which Nature 
does not do, upset Nature's distribution of brightness, etc. 
This is a phase of color quite the opposite to that which 
should interest us in our work-places. 
Fundamental Lighting Principles 
There are many lighting principles which we may well 
adopt from natural lighting outdoors on the safe assumption 
that we are best adapted to Nature's environment. Ob- 
viously, outdoors we have a compromise of conflicting condi- 
tions, but certainly man's egotism, admittedly and obviously 
gross as it is, cannot muster argument that the human being 
is not best suited to it or that this compromise is not the 
best suited to living things in general. There are obvious 
opportunities for improvement. For example, the great 
variation in daylight due to sky-conditions, and particularly 42 
LIGHT AND WORK 
to moving clouds, is not necessary. It adds charm to land- 
scapes, but it does not aid vision. On the other hand, out- 
doors it does not hamper vision. The great intensities of 
daylight compensate for the variation and make it unfelt, 
for there is usually enough daylight outdoors notwithstand- 
ing the variation. Coming indoors has placed limitations 
on daylight which are man's responsibilities, but in develop- 
ing an art of artificial lighting we should heed the great en- 
vironmental factors which are really fundamental lighting 
principles. Therefore, let us select from the foregoing dis- 
cussions what appear to be fundamental principles. We 
find in natural lighting conditions certain outstanding char- 
acteristics. The fourteen points are as follows: 
a. Great intensities of illumination averaging thou- 
sands of foot-candles throughout most of the 
day are characteristic of natural lighting. These 
values are literally a thousand times greater 
than mankind commonly provides by artificial 
means. It will be seen in Chapter VIII that the 
visual processes are divided into two regimes. 
The normal ones are operating under daylight 
intensities of illumination and the subnormal 
ones are operating at illumination intensities 
commonly found in artificial lighting. In other 
words, the eye may be said to be in " high " un- 
der the former conditions and in " low " under 
the latter conditions. 
b. Direct light from the sun is reduced in harshness 
by an appreciable percentage of light from a 
large area - the sky. Skylight is commonly 20 
per cent of the total light on a clear day. The 
light from this large area illuminates the 
shadows and not only reduces the harshness but 
also provides satisfactory visibility in them. DAYLIGHT OUTDOORS 
43 
Physiological laws indicate that as the level of 
illumination decreases even a greater percentage 
of light should reach the shadows. Outdoors 
generally there is plenty of light everywhere. 
This eliminates uncertain and faltering attitude, 
thereby making greater speed possible in work 
and in traffic with increased safety. 
c. Light coming solely from a large source such as 
an overcast sky is not as satisfactory as the 
combination of direct sunlight and diffused sky- 
light. Definite shadows are generally essential 
to satisfactory visibility and on overcast days 
the shadows are not well enough defined. Fur- 
thermore, comparatively shadowless illumination 
is characterless, uninteresting, and even depress- 
ing. 
d. The brightnesses of objects on an overcast day are 
not great enough relative to that of the sky for 
the best concentration of attention and the best 
visibility. The extremely high level of illumina- 
tion outdoors usually obscures this characteristic 
of overcast days but it is apparent to one who 
will search for it. It is particularly important 
under the lower levels of totally indirect arti- 
ficial lighting indoors. 
e. Outdoors an object casts only a single shadow. 
This is a fundamental principle of satisfactory 
visibility. It is too often ignored in lighting in- 
doors. There need not be only one light-source 
- window or artificial source - but the best 
visibility will be obtained where there is one that 
is predominant. 
f. The predominant light-sources are overhead and the 
eyes are so located and protected that they can 
avoid seeing the light-sources. In other words, 44 
LIGHT AND WORK 
the sun and most of the sky are generally out 
of the direct line or field of vision. 
g. The sun is so extremely bright that even careless 
and indifferent mankind can not bear to look 
at it directly. The brightness of the sky is of 
such a low value that at the levels of outdoor 
illumination it can be viewed with comfort. The 
highest brightnesses in the sky besides the sun 
are sunlit clouds (and white objects) and their 
brightnesses do not exceed 20 candles per square 
inch. It is well to aim at lower maximum bright- 
nesses indoors and certainly not to exceed this 
which can be taken as the maximum to which 
the eye is really adapted. 
h. Generally there are no excessive contrasts that are 
trying to the eyes. Those that would be trying, 
such as the sun and the adjacent sky or the 
image of the sun reflected from water and other 
" mirror " surfaces, are so great (and so few) 
that they are avoided. 
i. Nature's distribution of brightness is such that ob- 
jects in the foreground and general field of view 
are of much less brightness than those (sky and 
clouds) in the upper part of the field. This is 
a " natural brightness distribution " which it is 
well to retain indoors. 
j. The spectral composition of noon sunlight is such 
as to be approximately a true white. It is logi- 
cally the best quality of light for color discrimi- 
nation. It may be said to render colors properly 
for it does not appreciably " favor " any colors. 
k. North skylight in north latitudes is the least vari- 
able skylight available, both in intensity and in 
spectral character. For this reason it has be- 
come the standard for color-work, but it is not DAYLIGHT OUTDOORS 
45 
the best from the viewpoint of spectral charac- 
ter. It being decidedly bluish, it " favors " colors 
reflecting blue light such as violets, blues, blue- 
greens, and purples. Sight has been lost of the 
fact that north skylight was chosen primarily 
for steadiness and not primarily for spectral 
character or color-value. In establishing an 
artificial illuminant for color-work where steadi- 
ness can be easily attained noon-sunlight or a 
spectral character close to it which gives white 
light should be standardized. 
l. Daylight is a combination of sunlight and light 
from the blue sky. Inasmuch as the blue sky 
does not contribute over 20 per cent of the total 
light, average daylight does not depart radically 
from noon sunlight. Certainly this light is the 
best in spectral character for normal eyes be- 
cause of the eons of adaptation to it. This qual- 
ity of light should be more appealing to the eyes 
and to the entire psycho-physiological process 
of vision than light of any other spectral charac- 
ter, assuming all other lighting conditions are 
equal. Experiments and experience indicate 
that this is true so that the ideal artificial illu- 
minant should be one approaching the quality 
or spectral character of average daylight. In 
consideration of color-work a white light is de- 
sirable. Therefore, from all angles light of the 
quality of noon sunlight or of average daylight, 
which is approximately the same, is the ideal 
artificial as well as natural illuminant. 
m. The visual sense-organ is not only adapted to the 
visible radiation in daylight but has developed 
resistance against harm from the accompanying 
invisible radiations. Certainly an artificial il- 46 
LIGHT AND WORK 
luminant which emits radiations not found in 
solar radiation should be under suspicion until 
these other radiations are proved to be harmless. 
n. If there are radiations missing from the artificial 
illuminant that are present in the visible spec- 
trum of solar radiation it is not safe to assume 
that this is a satisfactory illuminant until ade- 
quate proof is available. Before and after cleaning the glass skylights. The photographic exposure 
was the same in the two cases. Chapter III 
DAYLIGHT INDOORS 
AS civilization advanced, activities were developed in- 
doors and, in the absence of powerful sources of arti- 
ficial light, it was necessary to bring daylight indoors. 
However, many handicaps were imposed upon daylight 
when mankind came indoors. Some of these are not ap- 
preciated at the present time because we have long accepted 
them through necessity. In the consideration of daylight 
indoors, the sky becomes the primary source of natural light. 
The sun cannot shine into all our windows and, furthermore, 
it often fails to shine at all. For this reason it cannot be 
depended upon in the design of natural lighting indoors. 
When it does shine indoors it is a mixed blessing. Direct 
sunlight is cheerful, but it often causes glare and visual dis- 
comfort amid the conditions indoors. 
Solar radiation has generally been considered valuable 
as a germicidal agent but this may be questioned after it 
passes through glass. Owing to the construction of build- 
ings most of our daylight openings are vertical windows in 
the walls. The result is very non-uniform distribution of 
light throughout the room. The daylight intensities of il- 
lumination indoors are very much less than outdoors and 
the variation, therefore, is more noticeable and unsatisfac- 
tory than it is outdoors. Indoor daylight not only varies 
with the variation of outdoor daylight but it depends upon 
exposure, glass area, location of openings above the ground, 
adjacent buildings and other obstructions to skylight, and 
to other factors, many of which are not under the control 
of the owner of the building or of the occupant of the room. 
47 48 
LIGHT AND WORK 
Finally, by bringing daylight indoors, it begins to cost 
us something. It is no longer furnished free as it is out- 
doors. Considering the matter from all viewpoints, day- 
light indoors has so many handicaps imposed that artificial 
lighting has become a formidable competitor from nearly 
every viewpoint. In fact, it will be shown that there are 
many conditions where artificial light is better and even 
cheaper than daylight indoors. 
Daylight Openings 
If we had to consider only the problem of admitting day- 
light into a room, this could be done simply and very well 
by means of a hole in the wall or roof. But such an open- 
ing would expose the room to weather, smoke, and dust and 
would also permit the escape of a large amount of heat. 
Nevertheless, many early buildings were daylighted in this 
manner before the general use of glass. These open " sky- 
lights " were common even where civilization reached such 
heights as in classic Greece. Some of the early temples 
were mysteriously lighted from " above." The evolution 
of building is a complex subject. Superstition and religion, 
as well as intelligence and available building materials, 
affected design and construction. Perhaps no factor was 
more influential than climate. For example, in Persia, a 
land of sunshine, open courts were commonplace in 450 b.c. 
The temples were open to the sky, for in their ceremonies 
light, the sun, the moon, and the elements were important. 
Clerestory windows appeared in early centuries in Assyrian 
temples. In Grecian and Roman temples, light was ad- 
mitted in many cases solely through doorways. In the 
sunny climes of the Mediterranean shores, close to the cradle 
of modern civilization, daylight openings could be relatively 
smaller than in the countries to the north. 
As the art of glass-making developed, window-openings 
could be extended in area. It is interesting to note how DAYLIGHT INDOORS 
49 
much greater the window areas became in northern Europe, 
where the intensities of daylight were less than in the sunny 
southern portion. Glass and fireplaces did much to diminish 
the handicaps of severe climate. The great Pantheon, 
which still stands as the most perfect preservation of ancient 
buildings in Rome, is lighted by a circular unglazed open- 
ing (27 feet in diameter and 143 feet above the floor) at 
the top of the hemispherical dome. 
As the complexity of civilization increased, buildings were 
constructed of a greater number of stories, so that the ver- 
tical skylight in the walls became more practicable from 
the viewpoint of construction. Therefore, the daylight 
opening most familiar to us is the vertical one located so 
we can see out of it. For the sake of simplicity these open- 
ings in the walls will be termed " windows." Daylight 
openings in the ceiling or roof will be termed " skylights " 
regardless of their slope. 
From the viewpoint of uniform distribution of light over 
the floor or other important horizontal plane, the overhead 
skylight is generally superior to so-called windows. Only 
on top floors or in single-story buildings are they particu- 
larly effective. Being overhead, they are out of the direct 
visual field. On the other hand, windows are often in the 
visual field and these are commonly sources of glare and vis- 
ual discomfort amid the relatively dark walls of interiors. 
From almost every angle, natural lighting indoors, by means 
of windows, is extremely unsatisfactory when one considers 
it in terms of outdoor daylight or of what can be done with 
artificial light. Where fairly high intensities of daylight 
can be obtained indoors, some of the unsatisfactoriness of 
non-uniformity of illumination is neutralized. However, 
natural lighting indoors in our cities is generally far from 
ideal in direction, distribution, steadiness, intensity of illu- 
mination, and duration of adequate illumination intensities. 
Windows, besides letting light in, "let vision out." It 50 
LIGHT AND WORK 
is a pleasant relief to look out of a window provided the 
outlook is at all desirable. It must be largely for this reason 
that windows extend so far down into the lower walls. The 
lower portion of low windows is not as effective in lighting 
a horizontal work-plane as the upper portion. In our 
crowded smoky cities it is often a problem to obtain enough 
glass area, so that nearly all the exterior wall is used for 
windows. However, there are conditions where the natural 
lighting would be better if only the upper area of the walls 
were used for windows. This reduces glare by eliminating 
sky-areas from the ordinary visual field and tends to keep 
the dominant direction of light from being too close to hori- 
zontal. Clerestory windows or openings high in the walls 
have excellent applications in interior lighting. 
There are many forms of overhead skylights designed 
chiefly to admit skylight without admitting, or at least dif- 
fusing, direct sunlight. The simplest is the flat (or nearly 
so) skylight. This utilizes direct sunlight as well as sky- 
light. If the glass is of such a character that it diffuses the 
sunlight, the lighting is improved for a sunny day but at 
a sacrifice of light on the overcast day when this loss of 
light can be least afforded. Any skylight glass which dif- 
fuses light sends light back outdoors as well as diffusing 
that which passes indoors. Skylight does not need to be 
diffused, hence, diffusing glass robs the indoors of some 
light that it would obtain if the glass were clear. There 
are clear glasses with ribs which spread the light. These 
glasses are more efficient transmitters of light if the ribbed 
side is toward the light-source but this is impracticable to 
the extent that greater accumulation of dirt becomes an 
economic factor. Prism glass is also available for diverting 
the direction of the light by refraction. All these have ex- 
cellent applications in daylighting as well as artificial light- 
ing but they must be used with understanding. 
Much study has been given to roof skylights,13 particu- DAYLIGHT INDOORS 
51 
larly for factories. Sometimes direct sunlight is desired 
but often the glass areas are deliberately faced to the north 
in order to admit only skylight. Furthermore, advantage 
is taken of light reflected from the roof areas, walls, etc. 
adjacent to the glass areas. These adjacent areas contrib- 
ute little light unless they possess reasonably high reflec- 
tion-factors. We have many designs of roofs for admitting 
daylight. The chief types are flat, monitor, saw-tooth, op- 
posed saw-tooth, and opposed monitor. These have various 
advantages and disadvantages depending upon the char- 
acter of the surroundings and of the work. They have 
value as ventilators in factories but they can be applied 
only to one-story buildings or only to the top floor. No 
factories today can operate on a regular full shift through- 
out the year without a supplementary artificial-lighting sys- 
tem. Hence, in consideration of such factors as the cost 
of elaborate daylighting equipment, the other costs charge- 
able to natural lighting, the undependability of daylight 
throughout the period of factory operation, the application 
of roof-skylights to only one floor, the non-uniformity of 
natural lighting from windows in large rooms, and the de- 
creasing cost of artificial light, it is reasonable to expect a 
growing interest in artificial lighting and a decreasing ex- 
penditure in elaborate daylighting equipment in many cases, 
particularly in congested districts. 
Intensity of Illumination 
Intensities of illumination indoors are of a different mag- 
nitude than outdoors. Of course, they vary considerably, 
but where intensities are measured in thousands of foot- 
candles outdoors, they are not generally measured even in 
hundreds indoors. Although it is difficult to devise an ac- 
curate statement for something so variable as daylight il- 
lumination indoors, it may be safely stated that the inten- 
sity of illumination seldom exceeds 100 foot-candles except- LIGHT AND WORK 
ing close to a window or skylight. Throughout rooms lighted 
by means of windows, the intensity of illumination varies 
enormously.14 Measurements show this range to be from 
almost zero to 100 foot-candles in many cases, but generally 
the average is less than io foot-candles. 
The foregoing values are for rooms lighted by windows 
on one side of the room from a sky whose skyline is fairly 
free from obstructions. In many places in our congested 
cities, daylight indoors is so feeble and of such a short dura- 
tion, that it is not worth even the extra heat-loss from the 
glass areas. The operation of artificial lighting in many 
stores and offices throughout the day is evidence of the in- 
adequacy of daylight indoors even during midday. Eco- 
nomic aspects are not immune from change any more than 
the other phases of civilization. Today many daylighting 
installations are yielding little or nothing on the investment 
represented. In many places money is spent on natural- 
lighting equipment where artificial-lighting equipment must 
provide lighting all day long. Such expenditure is largely 
wasted but this waste is still going on just because it has 
always been the custom to supply daylight openings. 
The low intensities of daylight illumination indoors are 
easily accounted for if we stop to think that direct sunlight 
is not a factor and that only a small portion of the entire 
sky is visible from a given point in a room. If on a clear 
day the entire sky outdoors contributes twenty per cent of 
the total light reaching a certain area on the earth's surface, 
the maximum skylight illumination of the area would be 
about 1500 to 2000 foot-candles in middle latitudes. This 
is based on the assumption that the area receives light from 
the entire sky - a hemispherical angle or a solid angle of 
180 degrees. If one thinks of the " pyramid of light " with 
the eyes at the apex and the skylight area seen through the 
window as a base, it will be obvious that this solid angle is 
but a small portion of a hemispherical angle. It is also easy 
52 DAYLIGHT INDOORS 
53 
to visualize the window in terms of an entire hemispherical 
dome of sky, the latter being suggested by the walls and 
ceiling of the room. Either method will convince one that 
the sky-area visible from most points in a room is generally 
only a few per cent of the total sky. For example, if only 
five per cent of the sky is visible, the intensity of illumina- 
tion at the point where the eye is would be 100 foot-candles 
on a clear day at noon when the sky is contributing 20 per 
cent of the entire 10,000 foot-candles outdoors. If there 
are several windows the total percentage of sky area visible 
from a point indoors would increase, other conditions being 
equal; however, waning daylight, city smoke, dense clouds, 
and other factors also must be taken into account. When 
all the factors are considered, it is easy to account for the 
low intensities of daylight illumination indoors. 
Although this book does not aim to discuss the details of 
design of lighting installations, it is so easy to gain an idea 
of the results from a patch of sky that it seems worth touch- 
ing upon. The illumination in foot-candles produced on a 
surface (perpendicular to the path of light directly from 
the source to the object) is found by dividing the candle- 
power of the source by the square of the distance (in feet) 
from the source to the surface. This " inverse-square " law 
does not hold accurately unless the distance between source 
and object is at least ten times the maximum dimension of 
the source. However, for approximations the distance can 
be as little as twice the maximum dimension of the source. 
In the case of a window the visible area of the sky is the 
actual source 15 but in effect it is the area as shown in Fig. 
6, of the window-opening " occupied " by sky as seen from 
the point in question. The candlepower of the source is 
easily found by multiplying the area in square inches of 
the portion of the window occupied by the sky, by the 
brightness of the sky in candles per square inch. This 
product is divided by the square of the distance in feet 54 
LIGHT AND WORK 
from the window to the point in question. It should be 
noted that this is the maximum illumination intensity at 
the point in question because the surface was considered to 
be perpendicular to the general direction of light from the 
source to it. If the surface is inclined to this perpendicular 
plane, the illumination intensity at the surface is diminished 
in proportion to the cosine of the angle of this inclination. 
Inasmuch as horizontal surfaces in a room are very much 
5ki| -area 
Surface. 
Fig. 6. Showing the effective sky-area at a surface at some distance from a 
window and also showing the general direction of light from the sky-area to 
the surface. 
inclined to the direction of the dominant light from win- 
dows in the walls (particularly low windows), the illumina- 
tion intensities of these surfaces are very much less than 
surfaces perpendicular to the general direction of the light 
from the sky-area to the surface. That is, they are much 
smaller than they would be if the surface " faced " the win- 
dow. Anyone interested in the least in the production and 
working conditions of even a few persons should have avail- 
able a " foot-candle meter " which is not expensive. 
The sky is not uniformly bright so that an average bright- 
ness must be used in estimates and computations. Clear 
skies are generally brighter near the horizon than near the zenith but the converse is generally true of cloudy skies.3 
The values of brightnesses for different kinds of sky given 
in Table III will be found quite useful. 
The intensities of illumination on a horizontal plane ob- 
tainable from overhead skylights are generally much greater 
than from windows. This is chiefly due to the facts that 
the light-sources are overhead and illuminate the horizontal 
plane directly, and there need be no floor-areas far distant 
from a skylight area. 
DAYLIGHT INDOORS 
55 
Reflection from Surroundings 
In the preceding discussion the lighting value of a day- 
light opening is based on the visible area of sky and its 
brightness. Earth-areas are generally of such a relatively 
low brightness compared with the sky that they need not 
be considered. This is usually true of most buildings and 
adjacent roofs which are so located that they could reflect 
considerable light into interiors if they possessed moderate 
or high reflection-factors. In many cases much improve- 
ment can be effected in daylight indoors by painting im- 
portant exterior areas white, although as already pointed 
out, first it should be determined if the result will justify 
the cost. The construction of buildings of permanent white 
or " light-colored " materials is due to a recognition of their 
contribution to daylight indoors in congested cities. How- 
ever, the lack of universal enforcement of the use of such 
material in cities indicates that there is no general feeling 
that the results are eminently worth while. Certainly arti- 
ficial light must still be used in many congested districts 
even where there are " white " buildings. 
An area of white wall receiving direct sunlight can be 
much brighter than the sky on a sunny day. It can readily 
have a brightness of io candles per square inch. This value 
is so much greater than the average of the sky that if ap- 
preciable portions of this area are visible from skylight 56 
LIGHT AND WORK 
areas, a great improvement in lighting on sunny days re- 
sults. However, on cloudy days when this additional light 
is particularly desirable, it is much less. A white surface 
illuminated by only a portion of the sky and no direct sun- 
light can never be as bright as the average brightness of 
the sky. In fact, it seldom is more than a small fraction 
of a candle per square inch on cloudy days. At best, a ver- 
tical surface can receive light from only one-half of the sky 
and inasmuch as the directions of light rays are inclined 
to this vertical surface, its maximum brightness possible 
on an overcast day, and also generally when it does not 
receive direct sunlight, is much less than one-half of the 
average brightness of the sky. 
Snow on the ground contributes much light indoors. This 
light is reflected upward against the ceiling and is diffused 
throughout the room. Snow often greatly increases the 
amount of natural light indoors and it is particularly wel- 
come in this respect in the season when daylight intensities 
are generally much smaller than in other seasons. Light- 
courts, having more than one side, are additionally effec- 
tive owing to the reflection of light from one surface to 
the other.16 This is particularly true when they are painted 
white. 
The reflection of light from floor, walls and ceiling in a 
room improves the lighting by diffusing the light and by 
increasing the illumination intensities on work-places. The 
reflection-factors of floors and walls are quite important in 
natural lighting from windows because much of the day- 
light entering windows falls on the floor and walls and 
reaches the work-places after various reflections. Thus, it 
is seen that natural lighting indoors is a very complicated 
matter and at best is far from ideal. 
Quality of Light 
Although this subject is discussed in detail in Chapters 
V and VI, a few remarks are not out of place at this point. DAYLIGHT INDOORS 
57 
Skylight being the primary light-source from a general in- 
door viewpoint its quality or spectral character is impor- 
tant in natural lighting indoors. It has already been pointed 
out that, in the northern hemisphere, north skylight is the 
most uniform in intensity and spectral character because 
the sun does not influence the north sky as much as it does 
the portions of the sky in the east, south and west which 
lie in or close to its path. This has attracted many workers, 
particularly those engaged in color-work, to north light. 
However, it is easily shown that light from the clear north 
sky is bluish white and therefore is not the best quality of 
light for color-work. It favors certain colors and this should 
not be a characteristic of a " standard " light. 
Light is modified by reflection from colored objects. For 
example, light reflected from green foliage is greenish; from 
red brick is reddish; and from yellow walls is yellowish. 
Such colored surfaces are often adjacent to daylight open- 
ings and, therefore, they add tinted light to the light 
direct from the sky. This amount is often quite appreci- 
able especially if the colored surfaces receive direct sun- 
light. This adds another variation to daylight whose un- 
certainty and particularly a lack of recognition of it have 
often caused difficulty in accurate color-work. Further- 
more, this vitiation of daylight is not constant throughout 
the year or even throughout the day. In temperate zones, 
we have green areas of vegetation in summer which under- 
go changes in the other seasons. The progress of the sun 
throughout a day is such that an adjacent red brick wall 
may be in sunlight during a certain period, thereby contrib- 
uting reddish light, and when this is in shadow this contri- 
bution is greatly reduced. At this time another surface of 
a different color may receive direct sunlight and thereby 
contribute its particular influence. Where accurate dis- 
crimination of color is involved these effects are annoying, 
inconvenient, confusing and often costly. Artificial day- 
light is the logical solution because it can be accurate and 58 
LIGHT AND WORK 
constant in spectral distribution and invariable in intensity 
- another desirable characteristic of lighting for such work. 
Colored surroundings in interiors also influence the color 
of the light reaching the work. Colored walls, floor, and 
various objects are, in general, more influential in this re- 
spect in daylighting from windows than in artificial lighting. 
In the latter case usually much of the light comes directly 
to the work from the luminaire and the nearly colorless 
ceiling. In natural lighting from windows a great percent- 
age of the light generally reaches the work by reflection 
from the interior walls, etc. Needless to state, rooms in 
which accurate color-work is done should be finished in 
neutral colors not only from a lighting viewpoint but also 
from a color-vision viewpoint. 
Hot-house Effect 
This effect is useful or undesirable depending upon the 
conditions and requirements. It is explained by a trapping 
of radiation. Glass is transparent to most of the solar 
radiation; that is, throughout a long range of wavelengths 
extending much further into the infra-red than shown in 
Fig. i. Therefore, these radiations pass through the glass 
and are absorbed to a great extent by the interior walls and 
various objects. The result is an increased temperature of 
these absorbers. They then radiate energy but, being of 
a relatively low temperature in comparison with the sun, 
this radiation is of very long wavelength. Glass is opaque 
to these rays of long wavelengths; hence, the temperature 
of the interior increases considerably until an equilibrium 
is reached when the rate of loss of energy from the walls, 
floor, ceiling, and glass area to the outdoors equals the rate 
of flow of energy indoors. The walls, roof, and other 
bounding areas play some part in this effect. They absorb 
the solar energy and re-radiate the energy in long wave- 
lengths as relatively low-temperature bodies, just as the ob- DAYLIGHT INDOORS 
59 
jects in the interior do. In re-radiating this energy in long 
wavelengths, some of it enters the interior from which it 
has the same difficulty in escaping as already explained. 
This " hot-house " effect is made use of in growing plants 
indoors. In other buildings it raises the temperature with 
a saving of coal in winter and often with an increase in dis- 
comfort in summer. Many experiments have been made 
with the hope of developing a skylight glass which would 
not admit the infra-red of solar radiation but at best only 
a minor advantage could accrue from the use of such a glass. 
A large percentage of solar radiation is in the visible region 
which could not be absorbed without a corresponding loss 
of light. Even a greater portion of sky-radiation is in the 
visible region. If a glass did absorb the portion of sky and 
sun radiation which was in the infra-red region, the heated 
glass would re-radiate most of the energy as radiation of 
very long wavelengths. Approximately one-half of this re- 
radiated energy would be emitted indoors by the glass and 
would contribute toward a rise in temperature. Therefore, 
the net advantage of a special glass of this sort would be 
slight. 
As seen in Fig. i, a layer of water 0.9 inch in thickness 
completely absorbs infra-red radiation of wavelengths longer 
than 1400 m|x (1.4 |i). Such a layer would 11 cool " direct 
solar radiation appreciably without reducing the amount of 
light only very slightly. A few inches of water would be 
still more absorbent of the infra-red. Some further cooling 
of the interior would result by increased conduction of heat 
from the glass to the water if they were in contact. How- 
ever, by absorbing all the infra-red much heating effect is 
still possible because the maximum energy in the solar spec- 
trum is in the visible region. The absorption property of 
water and the hot-house effect are useful in specific cases. 
It was pointed out in Chapter II, that water-vapor in the 
atmosphere absorbed considerable quantities of infra-red 60 
LIGHT AND WORK 
radiation. Water-vapor is somewhat more transparent to 
solar radiation than water. The amount of water-vapor in 
the atmosphere is generally equivalent to a layer of liquid 
water somewhat more than one-half inch in thickness. 
Hygienic Value of Solar Radiation 
It is well known that certain ultraviolet rays in solar 
radiation are germicidal. This has already been discussed 
for such radiation outdoors. The radiation which kills 
germs effectively is not efficiently transmitted by glass. In 
fact, the spectrum of solar radiation, ending abruptly at a 
wavelength of about 290 mp owing to absorption of the 
shorter waves, probably by the upper atmosphere, contains 
only a narrow range of wavelengths which are germicidal. 
These are almost completely absorbed by ordinary glasses 
as indicated in Fig. 1. Hence, solar radiation indoors after 
having passed through glass has been very much over-rated 
in hygienic value. Most glasses begin to absorb radiation 
considerably at 350 mp and become opaque to it at about 
310 mp, although there is a good deal of variation among 
glasses as to the wavelength at which they become opaque 
to ultraviolet radiation. However, ordinary window and 
skylight glass may be said to transmit little radiation that 
is effectively germicidal. In making this statement it is 
realized that many on first thought will consider it hereti- 
cal; nevertheless, the data available can not be interpreted 
otherwise. 
It has been well known to many investigators that germ- 
icidal rays were confined to a definite region of the spec- 
trum but apparently it has occurred to no one that ordinary 
glass filters out practically all these from solar radiation 
before it arrives indoors. Verhoeff and Bell10 found that 
radiation of longer wavelengths than 305 mp did not damage 
living cells to the point that their ability to repair the dam- 
age was overcome. Others have found that radiations DAYLIGHT INDOORS 
61 
longer than 310 mp in wavelength have no germicidal power. 
In fact, it has been found that the maximum of germicidal 
action is at 280 mp beyond the limit of solar radiation. The 
powerful effect of solar radiation is due to the great inten- 
sity of radiations relatively weak in germicidal effect. Some 
experiments with the quartz mercury arc showed that the 
germicidal action of this radiation was reduced to at least 
one thousandth of its original value when ordinary glass 
was interposed between the arc and the bacteria. These 
are just a few glimpses of the evidence which indicates that 
the germicidal value of solar radiation outdoors is greatly 
reduced and even practically eliminated when that radiation 
reaches indoors through glass. 
The therapeutics of radiation are not completely under- 
stood. There are many conflicting opinions but it seems 
safe to conclude that direct sunlight has some therapeutic 
value aside from germicidal action. These beneficial effects 
are not necessarily confined to the short wavelengths of 
ultraviolet radiation; if not, they would still be produced 
by solar radiation which had passed through glass. How- 
ever, the science of radiation-therapy is not sufficiently de- 
veloped, or at least there is still too much disagreement, to 
warrant an extensive discussion here. 
Fading of Materials 
This point is of passing interest excepting for those who 
are concerned with materials whose colors are fugitive. In 
general, only the light or radiation which is absorbed can 
be responsible for fading; however, it may be stated that 
ultraviolet, violet and blue rays are somewhat more active 
than radiation of other wavelengths. Moisture and heat 
generally hasten fading although radiation seems to be 
necessary. Direct sunlight is quite injurious to delicate 
colors owing to its great intensity and to its ultraviolet radi- 
ation. Even when filtered through ordinary glass it is still 62 
LIGHT AND WORK 
very injurious. Skylight is also very injurious but its lesser 
intensity causes it to escape with much less credit for fading 
than direct sunlight. Where delicate fugitive colors are 
concerned the solution is to exclude powerful daylight and 
to substitute the electric filament lamp with the " daylight " 
bulb. This bulb does not transmit very far into the ultra- 
violet. But the chief advantage of artificial daylight is 
that, with daylight excluded, a lower intensity than is usu- 
ally found in museums on bright days would be satisfactory 
for seeing and appraising the colors or the objects wearing 
them. Furthermore, artificial light can be turned on only 
when desired. This is not so easily accomplished with natu- 
ral lighting. This is a problem that has seriously confronted 
many persons. Its very obvious solution is by means of 
artificial lighting of proper quality and intensity and under 
thorough control. 
Cost of Daylight 
Daylight has been accepted for so long as a matter of 
course, that most persons are surprised to learn that it costs 
something. Of course, this is not true outdoors, but among 
the handicaps which man imposed upon natural lighting in 
bringing it indoors is the matter of cost.11 In general, natu- 
ral lighting indoors costs as much as artificial lighting and 
in some cases in crowded cities it costs much more. 
The chief items which contribute to the initial net cost 
of the natural lighting equipment are,- 
a. The difference in the cost of the building with and 
without windows and skylights. 
b. The ground-area sacrificed for light-courts and in 
other ways to obtain effective daylight openings. 
c. The cost of extra heating system to supply the ex- 
cess heat-losses from the glass areas over those 
from equal areas of walls, ceiling, etc. DAYLIGHT INDOORS 
63 
The chief items which contribute to the operating cost of 
natural lighting equipment are,- 
d. Interest on initial net investment. 
e. Depreciation of natural lighting equipment. 
f. Repairs. 
g. Cleaning glass. 
h. Depreciation of extra heating system. 
i. Extra fuel. 
There are a number of other items which may be charged 
to natural lighting but they are largely a matter of judg- 
ment so that they were omitted from consideration in ob- 
taining costs.11 From the viewpoint of this book the ac- 
tual cost figures of natural lighting indoors are not as in- 
teresting as a comparison of costs of natural and artificial 
lighting for the same room, house, or building. Quite a 
number of different cases were investigated and it was found 
that in nearly all cases natural lighting cost more than arti- 
ficial lighting. Artificial lighting, considerably above the 
present average and available at any time in constant quan- 
tity, costs less, in general, than variable, non-uniform natu- 
ral lighting whose availability depends upon the whims of 
nature. 
As a matter of interest to everyone it may be stated that 
the initial net cost of natural lighting in seven-room houses 
of different construction averaged about $600 and the total 
annual operating cost was about $100. These figures were 
closely approached by artificial lighting of a high standard 
considerably above average conditions. 
After having made the comparison on the basis of the 
items listed in a previous paragraph it may be of interest to 
present other items which could be charged to natural light- 
ing indoors. Some of these items are window-shades, spoil- 
age due to dust and leaks, fading of materials, sacrifice of 
floor space, and wall-areas, and cost of artificial lighting 64 
LIGHT AND WORK 
used during the natural lighting period when daylight is 
inadequate. 
Windowless Buildings 
In our congested cities daylight indoors is often inade- 
quate and the investment in natural lighting equipment is 
not paying dividends. In many factories, glass areas are 
so dirty that they are practically useless. The central por- 
tions of buildings now given over to light-courts could be 
utilized as floor space and artificially lighted and ventilated. 
On every hand there are conditions which would be eco- 
nomically improved by eliminating daylight and substitut- 
ing artificial light. The psychological effect of the absence 
of windows is well recognized, but in our congested cities we 
must work under inadequate natural light reinforced by 
artificial light, one of the least satisfactory lighting condi- 
tions. Many persons work all day long under a total ab- 
sence of daylight. Our complex civilization demands the 
services of thousands of night-workers who get along with- 
out daylight on their work. Artificial lighting installations 
are necessary in most places, so it would be an economic 
measure in many cases to divert the outlay for inadequate 
and undependable daylight to the rehabilitation of the arti- 
ficial lighting. In some cases, such as museums, the cost of 
natural lighting is so great and the result so unsatisfactory, 
that really wonderful artificial lighting could be supplied 
by combining both expenditures in artificial lighting. As 
it is, we often have inadequate natural lighting. By no 
means is this an advocation of a wholesale elimination of 
natural lighting. It is meant merely to call attention to 
the many cases where economics, if carefully studied, would 
show that natural lighting should be abandoned in favor of 
expenditure solely for adequate and proper artificial lighting. 
The question of ventilation naturally arises. The an- 
swer is that ventilation is a growing art which, like artifi- DAYLIGHT INDOORS 
65 
cial lighting, can stand firmly upon its own. Ventilation 
does not need to be an adjunct to natural lighting. It and 
heating are best achieved for all and everything concerned 
by not having to contend with the uncertainty introduced 
by the opening or closing of windows haphazardly. What 
makes these suggestions feasible, radical though they may 
appear, are the great developments of artificial lighting. 
Buildings are now being constructed or planned in some por- 
tions of which no daylight will be provided. 
We want daylight where we can obtain it adequately and 
at reasonable cost but we should recognize its cost and 
shortcomings. Our complex city life calls for many com- 
promises and also for much night-work. Almost every- 
where we need artificial light at night and in many places 
in the daytime. Those interested in building, particularly 
in congested districts, where land is expensive and the ef- 
fective sky is already greatly reduced in area, will do well 
to consider seriously the growing demand for adequate and 
proper lighting. Artificial-lighting possibilities are far 
ahead of those of natural lighting indoors in our crowded 
districts and to some extent everywhere. The economics 
of lighting indicate that we are at a turning point. Natural 
lighting in many places is bound to give way to artificial 
lighting because of the great handicaps imposed upon it by 
indoor civilization. 
Glasses for Daylight Openings 
Daylight openings as a rule should not be glazed with a 
diffusing glass such as an opal or " milk " glass. As already 
pointed out this highly diffusing glass diffuses about as 
much light outward as it does inward. If it is perfectly 
diffusing, it reflects more light outward than inward. If 
one holds a piece of highly diffusing glass in the path of 
light rays, it is readily seen that the glass is fully as bright, 
when viewed in the direction of the light rays as when 66 
LIGHT AND WORK 
viewed in a direction toward the light-source. In other 
words, a window glazed with this glass is fully as bright 
on the outside as on the inside. This means that about as 
much light is diffusely reflected outward as is diffusely 
transmitted inward. This loss of light is particularly unde- 
sirable when there is no direct sunlight. This is true to 
some extent, but not to the same degree, of sand-blasted 
or so-called " frosted " glass. Furthermore, this finely 
rough sand-blasted surface cannot be cleaned easily. 
For the foregoing reason, crystal glass is best for daylight 
openings but this glass does not need to be plain glass ex- 
cepting for windows that it is desirable to see out of. 
Ribbed and pebbled glasses are the best kinds of glass for 
breaking up the rays of direct sunlight without an appreci- 
able loss outwardly by diffusion. The irregularities are 
smooth and can be easily cleaned. Ribbed glass spreads 
the transmitted light in a direction perpendicular to the ribs. 
The pebbled glass spreads the light in an irregular manner. 
Both these types are generally very satisfactory where high 
transmission, ease of cleaning, and the breaking-up of direct 
solar rays are desirable characteristics. In Fig. 7, four typi- 
cal glasses - pebbled, coarse ribbed, fine ribbed, and wavy 
- are illustrated from prints made by placing them in con- 
tact with photographic paper and making prints by trans- 
mitted light. Their actual cross-sections are also shown. 
Illustrations made from actual photographs of the trans- 
mitted light are shown in Fig 8 for these four typical 
glasses. The transmission of light for the pebbled and for 
the wavy specimens are shown for two cases each. In one 
the rough surface is toward the light-source, in the other 
the smooth surface is toward the light-source. These illus- 
trate very well the manner in which the light rays are 
broken up. • - 
When light falls perpendicularly on clear plane glass 
about 4.5 per cent of the light is reflected from the first PEBBLED 
COARSE RIBBED 
FINE RIBBED 
WAVE RIBBED 
Fig. 7. Four typical crystal glasses having different surface characteristics 
as indicated by their respective cross-sections. PEBBLED 
C0AR5E RIBBED 
FINE RIBBED 
WAVY RIBBED 
Fig. 8. Showing the actual transmission of light through the four typical 
crystal glasses having surface contours as shown in Fig. 7. DAYLIGHT INDOORS 
67 
surface and about 4.5 per cent is reflected from the second 
surface. The percentage of light reflected increases with 
the angle of incidence as shown in Fig. 9 for three angles of 
incidence, nearly zero degrees, about 45 degrees, and about 
70 degrees. The principal reflections from the first and 
second surfaces are shown. 
The percentage of light reflected increases very gradually 
up to about 45 degrees incidence. It then increases rapidly, 
becoming 100 per cent at 90 degrees incidence. This is 
shown in Fig. 10 for one smooth clean surface and for the 
two clean surfaces of an ordinary plain and plane glass. 
100'1° 
4.4*1* 
!00'/» 
5.7°f» 
5a1> 
100$ 
18*'/- 
Glass 
67 y» 
9lV» 
68.9 */• 
Nearly 
perpendicular 
incidence 
Approximately 
45-degree incidence 
Approximately 
70-degree incidence 
Fig. 9. Showing the approximate percentages of light reflected by the first 
and second surfaces of clear plane glass at three different angles of incident 
light. The net transmissions are shown approximately, assuming no absorp- 
tion of light within the glass. 
The amount of reflected light varies with the optical prop- 
erty of glass, known as refractive index, but this variation 
is a minor factor in the consideration of ordinary glasses, 
because of the small range in refractive index. 
It is seen from the foregoing that no ordinary glass can 
transmit more than 91 per cent of the incident light and 
that at large angles of incidence (angle between the light 
ray and a line perpendicular to the surface) the transmis- 
sion is reduced very greatly. This is a very considerable 
loss in daylighting equipment. Furthermore, it should be 
obvious that the transmission-factor of even a plain glass 
is less for light reaching it from many angles than for a per- 68 
LIGHT AND WORK 
pendicular beam of light. For this reason in Table V the 
transmission-factors for various typical glasses are given for 
both a perpendicular beam and for light reaching the glass 
from a uniformly bright hemisphere, such as a uniform over- 
cast sky. The pebbled, ribbed, and wavy-ribbed glasses 
are the same as those illustrated in Figs. 7 and 8. Inas- 
much as the transmission-factor of a glass which is rough 
1.00 
0.90 
0.80 
0.70 
0.60 
0.50 
0.40 
0.30 
020 
0.10 
0 
C5econd surface 
First surface 
REFLECTION-FACTOR 
Glass 
One surface 
10 20 30 40 50 60 10 <30 90 
ANGLE OF INCIDENCE 
(DEGREES) 
Fig. io. Showing the percentage of incident light reflected by 
one and two surfaces of clear clean plane glass for various angles 
of incidence. 
on one side (sandblasted, pebbled or ribbed on one side) is 
appreciably less when the smooth side is toward the light- 
source than when the rough side is, the transmission-factor 
is given for both conditions. 
Other considerations being equal, ribbed, pebbled, and 
sandblasted glasses should be used with their rough sides 
toward the light-source, but unfortunately this is imprac- 
ticable in many cases. DAYLIGHT INDOORS 
69 
Transmission-factors of Typical Glasses for Direct and Diffused Light. 
TABLE V 
Specimen 
of glass 
Side toward 
light-source 
T ransmission-factor 
Perpendicular 
beam 
Hemispherical 
illumination 
Plane 
0.90 
O.8o 
Sandblasted 
rough 
.78 
•70 
u 
smooth 
•74 
•70 
Pebbled 
rough 
•85 
•75 
cc 
smooth 
•79 
•75 
Coarse ribbed 
rough 
•77 
.62 
M M 
smooth 
•52 
.62 
Fine ribbed 
rough 
.86 
•79 
U U 
smooth 
■79 
•79 
Wavy 
rough 
.88 
.82 
a 
smooth 
.86 
.82 
When a beam of light enters glass obliquely from air, 
the direction of its path is abruptly changed. The amount 
of this " rejraction " depends upon the refractive index of 
the glass. Advantage is taken of this property of glass to 
alter the course of light rays in the making of so-called 
prism glass. The refraction of light by prisms of glass is 
illustrated in Fig. n. Many designs of glass having pris- 
Fig. ii. Illustrating the refraction of light by a "prism" glass. 
matic forms are available. When such glasses are properly 
designed and installed, and also kept clean, an appreciable 
control of natural light entering interiors is obtained. They 
are usually employed for the purpose of gathering oblique 
light from the sky and sending it in a general horizontal 
direction into the parts of the room at some distance from 
windows. Unless kept quite clean prismatic glass loses its 
value. 70 
LIGHT AND WORK 
It is also possible to reflect all the light by means of 
glass prisms as shown in Fig. 12. After a ray of light 
enters glass it is totally reflected at a bounding plane if it 
is incident upon this plane at a sufficiently large angle. As 
seen in Fig. 12, the path of the light ray is altered 90 degrees 
I 
R 
I 
a 
b 
Fig. 12. Illustrating total reflection within glass prisms. 
in one case and is completely reversed in the other case. 
This total reflection within glass has many uses in the art 
of lighting. 
The foregoing is just a glimpse of a few important opti- 
cal principles which illustrate that an optical science is in- 
volved in lighting. The cleaning of window and skylight 
glasses is an important matter, but very much neglected, as 
shown in Table XV. Chapter IV 
ARTIFICIAL LIGHT 
IN preceding chapters, the salient characteristics of natural 
lighting have been discussed, and, before entering upon 
comparisons of it with artificial lighting, the development of 
artificial light logically demands discussion. Doubtless 
many readers are surprised that artificial lighting can be 
considered as a competitor of natural lighting. With due 
respect for natural light as a powerful environmental fac- 
tor and an important element in Nature's world, artificial 
light is now superior in many ways to natural light, handi- 
capped as it is by the shackles which man's artificial world 
has imposed upon it. If readers are surprised at what ap- 
pears rank heresy, or at least extreme temerity on the part 
of the author, in attempting to discuss the merits of artifi- 
cial lighting on the same level with natural lighting indoors, 
it must be due largely to unfamiliarity with the wonderful 
developments in the production and the utilization of arti- 
ficial light or with the cost and handicaps of natural light 
indoors. For the present let us inquire into the evolution 
of artificial light and particularly into developments of the 
present century. 
Fire 
Fire was the source of artificial light for countless cen- 
turies. The first primitive men who utilized fire will ever 
remain unknown, because the early utilization of fire for 
heat and light began long before mankind made permanent 
records of his achievements. In fact, the earliest use of 
fire was doubtless long before man devised weapons, vessels, 
71 72 
LIGHT AND WORK 
or tools, so that even these " records " are unavailable. 
From the viewpoint of this book we are not interested in 
speculation or even in the details of early light-sources. 
However, the use of fire for countless centuries, as a source 
of light as well as of heat, is an important factor. It is diffi- 
cult to imagine primitive men as being any more than beasts 
until fire was discovered and utilized. It is logical to be- 
lieve that the use of fire really established the earliest home 
and family circle.17 A cave remained a den until fire 
brought light, heat, and cheer and also protection from 
Nature's elements and perhaps from animals. Certainly by 
converting a cave into even a crude home, early primitive 
beings took the first important step up the ladder of 
civilization. 
For countless centuries, light from the flames of burning 
material was associated with cheer, comfort, and protection 
of home. This yellow-orange light could have had little 
influence as an environmental factor upon the physical and 
physiological processes of vision, for daylight was still in- 
comparably more powerful in this respect. But we cannot 
avoid the conclusion that light of this yellow-orange tint 
left its psychological imprint, for it was present in homes, 
however crude, during thousands of years when the founda- 
tions of our present civilization were being laid. Hence, 
it has become a symbol of comfort, of rest, of cheerfulness, 
of warmth, of home. Its feeble light could not replace the 
powerful light of the sun and sky and make night a period 
of activity comparable with day. Daylight continued to be 
the primary light for the growing activities of advancing 
civilization while firelight was the light of security and com- 
fort of the home and leisure. It is not surprising that the 
visual processes operate more efficiently and effectually un- 
der the high intensities of illumination with light of day- 
light quality. Indeed, it would be surprising if the feeble 
intensities of illumination with the yellowish light from ARTIFICIAL LIGHT 
73 
flames were effective from a visual viewpoint. Neverthe- 
less, no conception of light and its possibilities is complete 
without recognizing that, in some respects, the light of 
flames is deeply woven into the experience of evolving man- 
kind. This is discussed further in Chapters V and VI. 
Primitive Light-sources 
In those earliest centuries of primitive man, just as with 
primitive races today, the materials of nature were burned 
in their " natural " state. Later, as man developed obser- 
vation, memory, and reasoning sufficiently, he began to select 
materials which burned better and gave more light. Grad- 
ually he scoured nature for materials and then he thought 
of refining them. Vegetable and animal fats and oils were 
pressed into service of light-production. Vessels were 
shaped from rocks or clay to contain these fuels and a wick 
was devised. Receptacles of this sort are found in the 
earliest civilization of which we have a record. Even the 
origin of the candle, which is still a very convenient light- 
source, is obscure. It antedates the Chrisitan Era many 
centuries. It is a logical refinement of the principle of the 
early oil-lamp. It consists of a wick surrounded by hard 
fuel which is liquefied by the heat of the flame. It is little 
different in principle than the pith of a rush soaked in 
grease or oil. It is merely a refinement of the rush-light 
which still burns in the peasant homes of hinterlands. 
Oil-lamps 
For centuries upon centuries the light-sources were wood- 
fires, pine-knots, wood-splinters, oil-lamps, rush-lights, and 
candles. It seems incredible that civilization was to pro- 
gress for thousands of years until a very recent period be- 
fore these crude light-sources of exposed and flickering 
flames were to be radically improved. Receptacles and LIGHT AND WORK 
fuels were refined but it was not until 1783 that Argand, 
a French chemist, placed a chimney upon a lamp. This is 
said to have been an accident. He was heating a bottle 
over a flame and the bottom cracked and fell out. It was 
hot and he set it down suddenly, by accident, over the 
flame. The flickering flame became steady. Pondering 
over the matter and experimenting further resulted in his 
invention of the lamp-chimney. All these details of past 
developments seem simple to us now and we wonder why a 
civilization which had constructed the Pyramids, the Gre- 
cian temples, the Roman aqueducts or the Gothic churches 
of the Renaissance, did not invent the lamp-chimney. 
The seven seas were scoured for whales to furnish the 
whale oil for the most satisfactory light-source of Colonial 
days. The pools of petroleum in the earth were hidden 
from mankind until Drake tapped them in Pennsylvania in 
1853- 
74 
Gas Light-Sources 
We begin to see some experiments in the sixteenth cen- 
tury which supplied the knowledge leading to the manufac- 
ture of artificial gas from coal. Murdock distilled gas from 
coal, piped it into his house, and had gas lighting for the 
first time in 1794. 
This led to a period of gas-lighting by means of open 
flames until 1885 when Welsbach invented the mantle. It 
gave more light per unit of burned gas than was yielded by 
the open flame. This was due to a peculiar property of 
certain oxides such as ceria, thoria, and zirconia. If one 
will burn a piece of magnesium ribbon in the bunsen flame 
of a gas stove, a white ribbon of ash (magnesia) remains. 
Hold this in the bluish flame of the gas burner and it will 
glow brilliantly, with a brilliance quite out of proportion 
to its temperature. In doing this we shall have witnessed 
the principle of the gas-mantle and shall have repeated, ap- ARTIFICIAL LIGHT 
75 
proximately, the experience which led Welsbach to develop 
the gas-mantle. 
Electric Light-Sources 
Until 1800 practically nothing was known of electrical 
phenomena excepting that gained by rubbing a glass rod 
and producing a static charge. Some little work had been 
done along the line of stimulating electrically a leg of a 
dead frog, and noting that it twitched. No powerful source 
of electricity was available. Volta in 1800 created a sen- 
sation in the small and uncertain scientific world of that 
time, by announcing that a pile of plates, alternately zinc 
and copper, separated by means of wet cloths, was a source 
of appreciable current. This Volta's pile was improved and 
was the forerunner of the voltaic cell. Sir Humphrey Davy, 
by means of these " chemical " sources of electric current, 
extensively studied electric sparks and arcs and also ob- 
tained light by heating filaments of wire and of carbon. 
Here we see crude electric light-sources merely as scientific 
demonstrations. Their general use had to await, for more 
than a half century, the development of a more practicable 
- a mechanical - source of electricity. 
Organized scientific research was now just beginning. 
Faraday, a pupil of Davy's, thoroughly investigated electro- 
magnetics and the effects of cutting a magnetic field by 
means of an electrical conductor, such as a copper rod. He 
laid bare these laws between 1830 and 1840 and at the 
same time established the foundation upon which electric 
dynamos were developed a score of years later. Thus we 
see that all the tremendous developments of electricity have 
taken place in the past hundred years. 
When the efficient and adequate source of electrical en- 
ergy appeared about i860, the stage was perfectly set for 
the development of electric lighting. Not exactly the first, 
but at least the best known and most accomplished, actors 76 
LIGHT AND WORK 
to appear on this stage were Charles F. Brush with the elec- 
tric arc-lamp, in 1877, and Thomas A. Edison with the 
electric filament-lamp in 1879. They must be credited with 
the development of the entire lighting systems, including 
electric generator, distribution lines, electric light-sources. 
Now civilized man for the first time ceased to burn some- 
thing directly in obtaining artificial light. However, the 
idea of burning something became so deep-seated that we 
still speak of burning electric filament lamps when, as a 
matter of fact, the filaments are not " burned " at all in 
the ordinary sense of the word. 
From these electric lamps and the Welsbach gas-mantle 
which appeared soon after, lights were obtained which were 
whiter than the yellow-orange flames which had served man- 
kind throughout uncounted centuries. With the develop- 
ment of electrical science and art, various electric illumi- 
nants appeared. These became more and more efficient. 
The eye demands that the quality be within a certain range, 
as discussed in Chapters V and VI. Furthermore, light- 
sources must be practicable, safe, and of reasonable sizes, 
both physically and as to luminous output. Thus there have 
been several factors which have determined the superiority 
of a light-source for general use. 
Electric Arc-Lamps 
After the generation and distribution of electricity be- 
came highly practicable, various electric light-sources ap- 
peared. The gas-mantle arrived at an opportune time to 
keep gas-lighting for some time in the field against electric 
lighting. However, it is now giving way generally to elec- 
tric lighting and its future is not promising at present. The 
electric arc-lamp and the electric filament-lamp appeared 
at about the same time and, for many years, they were keen 
competitors. The arc-lamp reigned more or less supreme 
for a long time where light-sources of large luminous out- ARTIFICIAL LIGHT 
77 
put and efficiency of light-production were primary factors. 
On the other hand, in such places as the home where smaller 
luminous output per light-source was desirable and efficiency 
of light-production was of secondary importance, the elec- 
tric filament lamp was supreme. Both types of electric 
lamps were steadily improved. 
The carbon arc was enclosed in a globe. This enclosed 
arc was more steady and the carbons lasted longer than in 
the earlier open-arc. Next certain chemicals were intro- 
duced into the carbons with an increase in efficiency. Much 
of the light now was emitted by the arc flame; hence, these 
were called flame-arcs. The open 'and enclosed arc-lamps 
emitted light chiefly from the hot spot on the positive car- 
bon called the positive crater. Of the flame-arcs the yellow 
flame-arc was the most generally used. Finally came the 
magnetite arc or luminous arc consisting of a positive elec- 
trode of copper and a negative electrode which is an iron 
tube packed with iron and titanium oxides. This is the 
only arc-lamp (excepting the mercury-arc) which is play- 
ing a lighting role in this country. It is in use chiefly in 
high-intensity street-lighting but even here the electric fila- 
ment-lamp is doing most of the lighting. 
Mercury-arcs 
Perhaps the first mercury-arc was produced by Way in 
i860. He permitted a fine jet of mercury to fall from a 
reservoir into a receiving vessel, the reservoir and the re- 
ceiver being connected between the poles of a battery. Be- 
tween the drops of mercury arcs were formed. However, 
it remained for Peter Cooper Hewitt to develop the modern 
mercury-arc consisting of a transparent tube of glass or 
quartz from which the air is exhausted and in the ends of 
which electrodes are sealed. Liquid mercury covers the 
negative electrode. Current passing through the vaporized 
mercury makes it luminous. In fact, here we have the true 78 
LIGHT AND WORK 
arc. The light is lacking in red rays. In fact, the visible 
radiation is of very few wavelengths and, therefore, the ap- 
pearance of colors generally is very much distorted. The 
mercury-arc is used for lighting to some extent in the in- 
dustries and in photographic studios. It has many inter- 
esting special applications, particularly the quartz mercury- 
arc which is used for sterilization, therapy, and in many 
cases where powerful ultraviolet radiation is desired. 
Electric Filament-Lamps 
Edison's first lamps of this type and those of his contem- 
poraries possessed a filament of carbon and it was very in- 
efficient compared with present-day filament lamps. For 
twenty-five years improvements were made but it was not 
until 1906 that a radical increase in efficiency was achieved. 
Then, Dr. W. R. Whitney developed the metallized carbon 
filament with a surface so modified that the filament could 
be heated continuously to a much higher temperature than 
previously, with a consequent increase in efficiency of light- 
production. Incidentally, efficiency increases rapidly with 
increasing temperature, so that the scientist in light-produc- 
tion is on the lookout for materials or methods of treat- 
ment for filaments which will make it practicable to operate 
them at higher temperatures. Carbon does not melt until 
a very high temperature is reached, but its evaporation limits 
the operating temperature in a practicable filament lamp. 
Dr. Whitney's process retarded evaporation. 
Tantalum withstands a higher temperature than that at 
which the carbon filament could be operated so the tantalum 
lamp appeared. For the same reason the tungsten filament 
superseded the tantalum filament. The Nernst lamp should 
be mentioned in passing. It was a fine achievement but too 
complicated to compete with the tungsten lamp. 
All this time filaments were operated in a vacuum be- 
cause, even in an inert gas, heat was conducted rapidly from ARTIFICIAL LIGHT 
79 
a fine filament with a corresponding reduction in tempera- 
ture and in efficiency. After Coolidge succeeded in making 
tungsten ductile it could be drawn into fine filaments. This 
led Langmuir to make a helical coil of fine tungsten wire, 
thereby getting the equivalent of a filament of relatively 
large diameter. Having found that the cooling effect of a 
gas did not increase as rapidly as the light-output increased 
from the increasing area of filament, he produced the gas- 
filled lamp. It had long been known that a vacuum facili- 
tated the evaporation of the filament material and that the 
presence of gas pressure would greatly diminish the rate of 
evaporation. Thus it is seen that a combination of simple 
principles with the development of ductile tungsten pro- 
duced the gas-filled tungsten lamp. By diminishing the 
rate of evaporation the filament can be operated at higher 
temperatures with a resulting increase in efficiency. 
The tungsten filament lamp can be made of almost any 
desired luminous output from the tiniest spark to sources 
of such luminous intensities that there is little or no appli- 
cation for them. Tungsten lamps have been made as large 
as 30,000 watts and there is no reasonable limit to the size 
from a manufacturing standpoint. This extreme flexibility 
as to size, combined with the fact that the filament is sealed 
safely in a glass bulb, makes the tungsten lamp adaptable to 
a great variety of situations. 
Luminous Efficiency 
As already stated, luminous efficiency, expressed in lumens 
per watt, has steadily increased and in recent years has in- 
creased markedly. The average luminous efficiencies of 
electric filament lamps in use each year from 1905 to 1923 
are shown in Fig. 13. Improvement was very slight for a 
long period up to 1906. It then increases rapidly owing 
to new developments. It is the recency of the great in- 
crease in luminous efficiency that has found the user of light 80 
LIGHT AND WORK 
unprepared in attitude to take advantage of lighting now 
possible. Certainly we can afford to use much more light 
now than even a few years ago. It is this increase in the 
efficiency of light-production that makes it possible for arti- 
ficial light to compete with natural lighting indoors. 
IZ 
12 
AVERAGE LUMINOUS EFFICIENCIES OF LAMPS SOLD 
(lumens per watt) 
10- 
-10 
fl 
-a 
Showing influence of Mazda C 
(gas-filied tungsten) lamps 
6 
• 6 
4- 
4 
Shoiuinq inf uence of 
Mazda B(vacuum 
tungsten,) lamps 
Z 
-Z 
o 
1905 
1910 
1915 
I9Z0 
0 
Fig. 13. Average luminous efficiencies in lumens per watt of electric filament 
lamps in use each year from 1905 to 1923. 
YEAR 
Now, as never before, we can have adequate and proper 
lighting at a reasonable cost. It will be noted that the 
values in Fig. 13 are for average luminous efficiencies of 
filament lamps in use. The lamps of large wattage are gen- 
erally much more efficient than those of small wattage. In- 
asmuch as the electric filament lamp is almost universally 
used at the present time, more details of luminous efficien- 
cies in lumens * output per watt are of interest. These are 
presented in Fig. 14 for the various electric filament lamps. 
The efficiencies of the gas-filled tungsten lamps of various 
sizes are particularly interesting, so that these values are 
given for lamps as manufactured in 1923. It is seen that 
* A source of one candlepower in all directions emits 4?r lumens. ARTIFICIAL LIGHT 
81 
the luminous efficiency of a 1,000-watt gas-filled tungsten 
(Mazda C) lamp in 1923 is 7 times that of the carbon fila- 
ment lamp of 1905 and about 2.5 that of the vacuum tung- 
2C- 
24 
24 
20 
IS 
ie 
14 
iz 
10 
8 
6 
4 
2 
Untreated carbon highest for period 1579-1905 
Metallized carbon in isoe 
Tantalum in isos 
Osmium in isos 
750 watt 
WOO watt 
LUMINOUS EFFICIENCIES IN LUMENS PER WATT 
300 watt 
500 watt 
400 waft 
Motion-Picture projection lamp 
Tungsten (vacuum) in iso 6 
200 watt 
100 watt 
75 watt 
150 watt 
50 watt 
Mazda C (gas filled tungsten)a5of i9Z3 
This tqpeof lamp appeared first in 1914 
a 
Fig. 14. Luminous efficiencies of various electric filament lamps. 
sten filament lamp which appeared in 1906. The gas-filled 
tungsten filament lamp first appeared in practice in 1914. 
Cost of Artificial Light 
The attitude of the user of light toward the cost of this 
fundamental necessity is interesting. The householder com- 
plains of a cost for artificial lighting which does not exceed, 
on the average, $2.00 per month. Perhaps it is unfortunate 
that this bill comes regularly and clearly defined. Day- 
lighting costs fully as much, but its cost is so completely sub- 
merged that it is not even suspected of existence. The in- 
terest on " decorative " bric-a-brac within sight of the 
householder, as he or she sits in comfort under artificial light- 
ing in the evening, often greatly exceeds the lighting cost. 
The cost of artificial lighting for the manufacturer and the 
merchant also glares at them in the form of a bill, but they 
little realize what that artificial lighting does for them or LIGHT AND WORK 
82 
what great costs are piled up by dark corners or inadequate 
general lighting. These advantages are discussed in other 
chapters. 
It is interesting to look backward over a century and see 
how artificial light has steadily decreased in cost.17 In Fig. 
100 
90- 
ao 
70 
60 
50 
40 
30 
20 
RELATIVE COST Of ARTIFICIAL LIGHT 
Tungsten filament (vacuum) lamp isoe-iszs 
Tungsten filament lamp I9H-1923 
Metallized carbon filament lamp i9O6-/5>io 
Kerosene lamp i650~ 1895 
Carbon filament lamp ia?5-i905 
Gas-flame 1665-1895 
Talloui Candle isoo-1675 
Whale-oil lamp I800~i650 
Gas-mantle i890-i9Z3 
10 
0 
Fig. 15. Relative costs of artificial light during various periods of the past 
century. 
i5 approximate relative costs of artificial light are pre- 
sented for various illuminants during the periods of their 
extensive use. It is seen that the cost of artificial light is 
now only one or two percent of its cost less than a cen- 
tury ago. This is a very wonderful achievement which has 
greatly extended the possibilities of artificial light and has 
made it a formidable competitor of natural light indoors in ARTIFICIAL LIGHT 
83 
every respect. The items which contribute toward the total 
cost of daylight indoors have generally increased in cost 
throughout the past century. Similar influences have been 
retarding the decrease in cost of artificial light but the scien- 
tific developments have so greatly increased the efficiency 
of light-production, that the net result has been a tremen- 
dous decrease in the cost of artificial light. 
The great economic value of scientific developments is 
scarcely realized by the public. In fact, it is difficult to 
MO 
120 
Food 
100 
RELATIVE COST 
60 
jjgnt 
60 
40 
20 
0 
1894 1696 1698 1900 I90Z 1904 1906 1906 1910 I9IZ 1914 
YEAR 
Fig. 16. The decreasing cost of electric lighting and the increasing cost of 
food for the twenty years preceding the World War. The war period is left 
out of consideration, owing to the abnormal conditions; however, it is interest- 
ing to note that the cost of electric lighting since 1914 has steadily decreased 
while the cost of food has greatly increased. 
illustrate this value in dollars. This is equally true of light- 
ing because there is so much value in increased safety, pro- 
duction, and happiness that can scarcely be appraised in 
terms of money. However, the relative decrease in cost 
of electric light compared with the relative increase in the 
cost of food as shown in Fig. 16 is elucidating, because the 84 
LIGHT AND WORK 
cost of food is at least a rough gauge of wages and the cost 
of necessities for all of us who work. The period was chosen 
between 1894 and 1914, the war having introduced such an 
economic disturbance that it seems best to stop at 1914. 
It may be stated, however, that electric lighting has not 
generally increased in cost since 1914 and everyone knows 
that the cost of food increased very much. The great de- 
crease in the cost of electric lighting during the twenty years 
preceding the war, while the cost of food and of many other 
things increased steadily, is striking testimony of the eco- 
nomic value of scientific research. It is evidence of the 
benefits reaped by the consumer through the expenditures 
of resources by manufacturers in the development of their 
products. The cost of artificial light has now reached a 
point where adequate and proper artificial light can be ob- 
tained at a cost equivalent to a very small percentage of 
the payroll of workers for which it is supplied. The in- 
creased benefits of good lighting over those of poor lighting 
are many times the additional cost. In fact, good lighting 
can often be obtained for less than the cost of poor lighting. 
In such cases it is a matter of proper knowledge applied to 
the design. 
The United States leads other countries in the use of light, 
but lighting even in this country is far below the standard 
that it should be. It is true that the average amount of 
artificial light used per inhabitant is increasing, but this in- 
crease is approximately in proportion to the decreasing cost 
of light. In Fig. 17 the approximate average lumens of 
artificial light used per inhabitant per year in this country 
are shown for a period of years from 1912 to 1923. Dur- 
ing this period most of the artificial lighting has been done 
by electric filament lamps. Gas-lighting has relatively de- 
creased and lighting by means of electric arcs has fallen to 
a very low percentage. The population of this country has 
been increasing at a uniform rate since 1900 and the lumens ARTIFICIAL LIGHT 
85 
output of electric filament lamps has been increasing also 
at an approximately (but much greater) uniform rate. 
Making estimates of the relative decrease in lighting with 
gas and with electric arcs, the rate of increase of consump- 
1400 
1200 
1000 
800 
600 
-400 
ZOO 
c 
LUMENS 
I9IZ 1914 1916 1916 I9Z0 I9ZZ 
Fig. 17. The approximate average lumens of light used m the United States 
per inhabitant each year from 1912 to 1923. 
YEAR 
tion of artificial light per inhabitant was found as shown in 
Fig. 17. It is seen that the rate for the period shown is 
only approximate but it can not be far from correct. The 
rate will become much greater as the user of light comes 
to realize the importance of adequate and proper lighting. Chapter V 
ILLUMINANTS AND COLOR 
ANY lighting problem or condition can be separated into 
three parts, namely, quality of light, intensity of illu- 
mination, and distribution of light. Intensity of illumina- 
tion is quantity of light per unit of area. Distribution in- 
volves direction and diffusion of light. Quality or spectral 
character is solely a characteristic of the light and the two 
other factors are characteristics of lighting. Lights can 
differ only in respect to quality because this is the only 
fundamental or inherent characteristic of light. Quality 
of light is intimately related to the color of light. A light 
of a specific quality is always of a certain color, but lights 
of the same color are not necessarily of the same quality. 
This is treated in detail elsewhere,1 so it will be discussed 
here with brevity. It should be noted that the visual proc- 
ess is a " synthetic " process. The eye may receive visible 
radiations of all wavelengths from the sun, for example, 
by reflection from a colorless surface, and, although these 
radiations of different wavelength cause various hue sensa- 
tions (violet, blue, green, yellow, orange, red and interme- 
diate hues), the combined sensation is white. If we mix 
a blue-green light and a red light, of just the right colors 
and in proper proportions, a colorless gray surface will ap- 
pear colorless when illuminated by this special mixture. 
These two white lights (the sunlight and the special mixture 
respectively) would be far different in quality or spectral 
character. This is discussed later and is introduced here 
only to aid those unfamiliar with the meaning of quality or 
spectral character of light. 
This is a phase of light whose importance and effects are 
less obvious to the average person than intensity of illumi- 
86 ILLUMINANTS AND COLOR 
87 
nation. Even many lighting engineers are less familiar with 
this aspect than with intensity and distribution. The rea- 
son in both cases is because it is less evident than intensity 
of illumination or distribution of light. Nevertheless, qual- 
ity or spectral character of light is very important, not only 
in everyday lighting but also in the scientific and artistic 
fields of light and vision. By quality of light is meant the 
relative amounts of energy of various wavelengths present 
in the visible radiation of an illuminant, or in other words, 
the spectral distribution of energy in the visible spectrum. 
By extension, the term may be applied to the ultraviolet and 
infra-red regions. 
This spectral distribution of energy of various wave- 
lengths is something we do not see. If we are familiar with 
certain scientific aspects we can often judge the quality of 
light by its color. It must be determined by means of a 
special instrument known as a spectrophotometer.1 The 
light to be studied is dispersed (usually by means of a glass 
prism) into its spectrum; that is, the radiations of various 
wavelengths are spread out into a band which we see colored 
progressively from violet to red as indicated in Fig. i. The 
same is done for a standard light, whose spectral distribu- 
tion is known. Then the intensity of radiation of any 
wavelength (or hue) can be measured in terms of the in- 
tensity of radiation of the same wavelength (or hue) by 
the usual photometric procedure. Progressing throughout 
the spectrum we obtain the spectral distribution of energy 
of the light under consideration. Thus we determine some- 
thing we cannot see directly; however, if we are observant, 
we do see the effect of spectral character of an illuminant 
upon colored objects. 
Spectral Character of Illuminants 
The distribution of energy of different wavelengths 
throughout the visible region is shown in Fig. 18 for sun- 88 
LIGHT AND WORK 
light at different times of the day. At noon it is seen that 
the maximum energy is in the middle of the visible spec- 
trum- the yellow-green. In Fig. i and Table I, it is 
pointed out that this is the region of maximum sensibility 
of the eye. It was also suggested that this is one of the 
imprints of daylight as an environmental factor during the 
evolution of man and his visual sense-organ. In fact, it 
is a most natural consequence of adaptation that the eye 
Ye I lour- 
120 
Violet Blue Green Green Yellcu; Orange Red 
-floon sun 
100 
80 
RELATIVE ENERGY 
60 
40 
zo 
o. 
'400 
500 
WAVELENGTH (m/i) 
MO 
100 
Fig. 18. The spectral character (spectral distribution of ra- 
diant energy) of the visible radiation reaching the earth from 
the sun at different altitudes or periods of the day. It is seen 
that the sun appears reddish toward sunset owing to the pre- 
dominance of yellow, orange and red rays. 
should be most sensitive to those wavelengths of radiation 
in daylight that are the most plentiful. In Fig. i is shown 
the evaluation of radiation of the various wavelengths by 
the eye; that is, the relative luminosities produced by equal 
amounts of energy throughout the range of wavelengths 
which causes the sensation of light. The various sensations 
are of different hues and brightnesses as shown by this 
spectral luminosity curve. In Fig. 18 only the relative 
amounts of energy of different wavelengths are shown for ILLUMINANTS AND COLOR 
89 
the different sunlights. This energy has not been evaluated 
as light by the eye. In fact, it is best to show the spectral 
energy distribution because it is not always the eye that 
we are concerned with. The photographic emulsion or any 
other photochemical process might be of interest, and its 
evaluation of the radiations of various wavelengths differs 
in general from that by the eye. Therefore, the spectral 
character of the radiation is best shown in terms of relative 
energy. 
As the sun leaves the meridian and decreases in altitude, 
sunlight must pass through increasing thicknesses of the 
atmosphere and amounts of smoke, dust, water, and ice par- 
ticles that are suspended in it. Owing to the selective 
transmission of the atmosphere, sunlight reaching the earth 
grows yellower toward sunset (or sunrise). This is well 
shown in Fig. 18 by the lower curves. Finally, the sun 
may appear quite red. Its spectral character when nearly 
setting is shown by the lowest curve. Thus these data not 
only show the change in the spectral character of sunlight 
with altitude of the sun but they illustrate very well what 
is meant by spectral character. 
In Fig. 19 the spectral distributions of energy in the 
visible region are shown for several illuminants 1 of interest 
from the viewpoint of this book. The spectral character 
of noon sunlight is again shown for the sake of comparison 
and because it is approximately white light. Skylight from 
a clear blue sky is seen to have relatively more radiation 
of the shorter visible wavelengths (violet, blue, green) than 
sunlight. The spectral character of the light from a candle- 
flame is seen to be quite the converse of blue sky; that is, 
it has relatively more radiation of longer wavelengths (yel- 
low, orange, red). It is much yellower than noon sunlight. 
The relative abundance of the yellow, orange, and red radi- 
ations in the light of the candle-flame makes it a yellow- 
orange tint instead of a colorless white light. It is common LIGHT AND WORK 
90 
experience that light from the carbon filament lamp is yel- 
lower than that from a tungsten (vacuum) lamp and that 
this in turn is yellower than light from a tungsten (gas- 
Violet 
Blue 
Green 
Yel Iouj 
Orange 
Red 
260 
260 
240 
220 
ZOO 
ISO 
160 
140 
120 
100 
80 
60 
40 
20 
0 
400 450 500 550 600 650 700 
tMez-z 1 
/A 
B, 
RELATIVE ENERGY 
C, 
D 
E 
F 
M 
M 
G 
M 
M 
M 
WAVELENGTH Cm/i) 
Fig. 19. The spectral distribution of visible radiation in various illuminants 
A Candle flame 
B Carbon filament 
E Mazda daylight lamp 
F Noon sun 
G Blue sky 
M Mercury arc 
C Vacuum tungsten filament 
(8 lumens per watt) 
D Gas-filled tungsten filament 
(22 lumens per watt) 
filled) lamp. These colors are matters of temperature of 
the filaments and the color of light emitted by virtue of ILLUMINANTS AND COLOR 
temperature of the solid radiator (filament, arc-crater, etc.) 
is an approximate indication of the quality of the spectral 
character of the light. However, this is not true of glow- 
ing gases or vapors such as the arc of an arc-lamp or the 
vapor of a mercury-arc. 
Perhaps this may be made clear by means of a tungsten 
wire heated by electric current. If a rheostat be placed in 
series with the wire and the room darkened, some current 
can pass through the wire without heating it to a glowing 
temperature. If we feel of the wire we will find it becomes 
warm when sufficient current flows but it does not give off 
light. If the resistance in series is decreased, more current 
flows and the wire will become hot. Still no light is given 
off. All this time radiation is being emitted by the wire 
but it is all of long wavelengths in the infra-red region, in- 
capable of causing the sensation of light and color. As we 
increase the current the wire finally begins to glow a dull 
red. Now some of the long-wave visible radiation is being 
emitted but the amount of infra-red radiation has also in- 
creased. As the current is further increased the wire glows 
brighter and appears orange in color. More visible radia- 
tion is being emitted and a greater percentage of this is of 
shorter wavelengths. In other words, as the temperature 
of this wire is increased, relatively more radiation of shorter 
and shorter wavelengths is being emitted and in greater and 
greater percentages of the total radiation. Finally, if we 
could heat the wire to a temperature of the sun we would 
have a white light of approximately the same spectral char- 
acter of noon-sunlight as shown in Fig. 19. 
All this time the efficiency of the wire as a light-source 
increased as we increased its temperature. When it did not 
glow at all its efficiency as a light-source was zero and the 
electrical energy put into it was wasted if our object was 
to obtain light. The highest luminous efficiency would be 
reached when the temperature was approximately that of 
91 92 
LIGHT AND WORK 
the sun and from then on the luminous efficiency would de- 
crease as the temperature was increased. This is because 
at about the temperature of the sun the tungsten wire would 
be emitting the maximum amount of energy in the visible 
region. Incidentally, it is interesting to note that the tem- 
perature of the sun is approximately the most effective 
temperature for producing light by means of bodies which 
emit light by pure temperature radiation. In this connec- 
tion it should be noted that the range of wavelengths of 
radiation to which the eye responds is the result of adapta- 
tion and the result naturally is a perfect adaptation of the 
sense-organ to solar radiation. 
In the foregoing experiment, we have witnessed the 
method of light-production which is now generally used. 
The chief aim in this field of scientific research is to operate 
filaments at the highest temperatures practicable. No 
solids have yet been found or developed which can be 
heated to temperatures approximating that of the sun and 
still remain in the state practicable for light-sources. The 
temperature of the incandescent solids used as light-sources 
has steadily increased and this has been the chief factor in 
the increase in luminous efficiency.17 The temperature in 
Fahrenheit and corresponding luminous efficiencies of a few 
of these light-sources are in Table VI. The lumens per 
watt corresponding to the sun's temperature as it appears 
from the earth is for white light emitted by a solid radiator 
like carbon if it could be operated at the sun's apparent 
temperature. 
TABLE VI 
Approximate Temperatures and Luminous Efficiencies of a Few Light-Sources 
Temperature 
Lumens per watt 
Untreated carbon filament 
3300 deg. F. 
3° 
Metallized carbon filament 
3400 
4-5 
Tungsten (vacuum) filament 
37oo 
8.0 
Tungsten (gas-filled) filament up to 
53oo 
10 to 26 
Sun as it appears at noon 
8900 
(100) ILLUMINANTS AND COLOR 
93 
It is not the intention to discuss spectral character here 
in detail; those who wish to go deeply into the subject will 
find it adequately treated elsewhere.1 However, it is of 
interest to touch upon luminous gases and vapors. The pre- 
ceding discussion was confined to solids which obey certain 
laws approximately and emit a continuous spectrum; that 
is, they emit radiations of all wavelengths between wide 
limits. In gases and vapors there is more freedom of the 
atomic radiators and when such materials as gases and solids 
are excited electrically, they emit discontinuous spectra 
consisting of lines and bands. Each gas or vapor emits a 
discontinuous spectrum characteristic of it and it alone. 
The individual spectrum is just as characteristic of the par- 
ticular gas or vapor as a finger-print is of a human being. 
This forms the basis of spectrum analysis. The scientist 
can recognize the spectral lines of an element when it is 
emitting radiant energy as a luminous gas or vapor. He 
finds these lines in the spectrum of a star and can tell what 
elements exist there in a gaseous or vaporous state. 
A number of light-sources have been developed in which 
gas or vapor are excited electrically. The Moore tube con- 
taining nitrogen was used to some extent years ago. The 
Moore carbon-dioxide tube emitted a white light and has 
been used to some extent for color-work. Some neon and 
helium tubes are also in use. Of all these so far the mer- 
cury-arc has been used the most extensively. Its visible 
spectrum consists chiefly of four lines, namely, violet, blue, 
green, and yellow. The latter is really a double line. These 
four spectral lines are designated in Fig. 19 by the letter 
M. The relative amounts of energy of these four wave- 
lengths are represented approximately by the lengths in the 
illustration, but owing to the greater luminosity of radiation 
in the mid-spectrum region (see Fig. 1) about 95 per cent 
of the total light from the mercury-arc is due to the green 
and yellow radiations, the two lines. Such a light from LIGHT AND WORK 
94 
which most of the spectral colors are absent causes almost 
everything to appear unnatural. Other phases of the light 
of the mercury-arc are discussed later. 
Natural Illuminants 
Daylight and its components - sunlight and skylight - 
have been emphasized as the natural illuminants. Of these, 
noon sunlight is the more powerful and represents fairly 
well the average quality of light throughout most of the 
day and year. It may be safely assumed to have been the 
environmental illuminant to which living things and their 
sense-organs became adapted. It, then, is the fundamental 
natural illuminant. But we should not overlook another 
quality of light - that from fire and other common flames 
- which has been associated with mankind throughout the 
long period he has struggled up the ladder of civilization. 
Viewed in this manner, the light from flames or flame-tinted 
light is a secondary natural illuminant. But in considering 
this, as has been done in Chapter IV, we must do so in con- 
nection with man's activity and its environment. Dur- 
ing the countless centuries of civilized progress from the 
most primitive state of man, daylight was the period of 
activity and firelight was associated with home and recrea- 
tion. This psychological aspect is important as shown in 
Chapter VI. 
Thus we see on one end of the scale of natural illuminants 
the yellow-orange light of flames and at the other end blue 
skylight. Somewhere between is noon sunlight. All these 
natural illuminants are now duplicated by man so that we 
may have them where needed or desired. They are ob- 
tained by filtering the light from tungsten filaments through 
specially developed filters of permanent color. The desired 
quality of light is obtained in this manner by a loss of light, 
but the high-efficiency tungsten lamps make this method 
economically possible. Where light of the quality of sky- ILLUMINANTS AND COLOR 
95 
light is needed and skylight is not available, artificial 
skylight over a small area can be obtained by properly 
filtering the light from a gas-filled tungsten filament lamp. 
Surely this small cost for an accurate substitute for day- 
light is justifiable. Furthermore, if the unfiltered light from 
the tungsten lamp will not answer the purpose we have the 
case of subtracting some of the rays and increasing the 
value of the remaining light. Artificial daylight devices 
are available which produce any quality of light from noon 
sunlight to clear north skylight. 
Light of a flame tint is now produced by properly filter- 
ing the light from a tungsten filament usually by coloring 
the lamp-bulb. Between the natural flames and the noon 
sun we have the tungsten filament lamps. One of these, 
the so-called tungsten " daylight " lamp, is an approximate 
daylight in quality. It is quite satisfactory for general 
light and it greatly facilitates the discrimination of colors. 
For very accurate color-work some artificial-daylight equip- 
ments are very satisfactory but these are not ordinarily 
used for illuminating extensive areas. However, in many 
places where persons work, and particularly when artificial 
and natural light are mixed, the tungsten daylight lamp for 
general lighting is quite satisfactory. It aids color-discrimi- 
nation and is more natural than the yellower illuminants 
for everyday work. 
In order to aid in visualizing the relations of the various 
natural and artificial illuminants to each other, they are 
listed in Table VII. Beginning with blue sky at the top we 
progress downward toward white (noon sunlight) and 
thence through yellowish tints of glowing solids to ordinary 
flames. On the right are listed the artificial illuminants 
which now cover the entire scale of natural illuminants 
although some of the natural illuminants are produced 
artificially by carefully prepared colored glasses or super- 
ficial coatings. 96 
LIGHT AND WORK 
TABLE VII 
Natural and Artificial Illuminants Arranged in Relation to Each Other along 
the Natural Scale of Spectral Character from Flames to Blue Sky. 
NATURAL 
ARTIFICIAL 
BLUI5HTINT 
QI lie- CtZYZ a r-i-r-l c-< l A 1 1 z-. .-r- ..k..-• 
UUUL I 
1 J 1 1 U IML OMLKini UNH 
( filtered light) 
Combinations of sunlight 
and daylight 
Any quality of daylight is ob- 
tainable bq filtering light from 
tungsten fi laments 
WH 
wnnm q11m - 
ITE 
N U U N j U N 
Ak 1 II ICIAL jUNLIUn I UNI i 
(Filtered light) 
TUNGSTEN DAYLIGHT LAMP 
(Filtered light) 
Glowing solids might 
be termed tertiary 
natural illuminants 
GAS-FILLED TUNGSTEN LAMP 
VACUUM TUNGSTEN LAMP 
TANTALUM LAMP 
METALLIZED CARBON FILAMENT 
UNTREATED CARBON FILAMENT 
Light from ordinary 
flames is a secondary 
natural illuminant 
(Filtered light) 
n AMF_TlMTPn I AMH 
1 LMTIL 1 LMI'IL IH1ILU LMI'IF 
YELLOW-ORANGE TINT ILLUMINANTS AND COLOR 
97 
What is Color? 
The quality or spectral character of light has already 
been defined. This term is also applicable to any colored 
light whether emitted by a light-source or reflected by a 
colored surface. Then, in general, all lights have a spectral 
character. In fact, this is the fundamental physical founda- 
tion of color. And just as the eye alone is incapable of de- 
termining the spectral character of light, so it is incapable 
of analyzing a color into its components. For example, a 
perfect yellow can be made by mixing red and green lights 
and this yellow will appear quite the same as an ordinary 
yellow. Not recognizing that the eye is synthetic as com- 
pared with the analytic ear, is the cause of many errors 
and lack of understanding of the relation of light to the 
appearance of colors. 
In order to gain a fundamental understanding of color it 
must be recognized that the color is really in the light. 
The pigment or dye is merely a selective absorber. It ab- 
sorbs radiations of certain wavelengths and reflects others. 
For example, vermilion has the physical property of absorb- 
ing most of the visible radiations and of reflecting the red 
rays. But it, like most colored media, reflects radiation of 
a certain range of wavelengths. It reflects, besides red, 
orange and some yellow but the integral sensation caused 
by these rays is what we loosely term red. It should also 
be noted that we can not assume that it reflects no invisible 
radiations. In fact, most colored media also selectively 
absorb and reflect the infra-red radiations but this is only 
of secondary importance to us here. 
This vermilion pigment appears red only when there are 
red rays in the illuminant for otherwise it would reflect no 
red light. This is shown by viewing it under light from the 
mercury arc. It is now almost black. It does not appear 
red now because there are no red rays in mercury-arc light 
as shown in Fig. 19. Our lips are bluish black under this 98 
LIGHT AND WORK 
light because they reflect some blue and violet and a great 
deal of red. If there is no red light in the illuminant but 
violet and blue rays are present the lips, of course, appear 
like those of a youngster who has just eaten blueberry pie. 
This accounts for the unnatural appearance of our surround- 
ings and of our ghastly appearance under the light from a 
mercury arc. This same process of reasoning can be applied 
to any colors.1 
If we illuminate two sets of a number of colors, for ex- 
ample, tints of violet, blue, green, yellow, orange and red, 
one by skylight and the other by ordinary artificial light, 
we shall see that the violet, blue and green are more brilliant 
and colorful under the skylight than under the ordinary 
artificial light. The converse is true of the yellow, orange 
and red. A yellow illuminant brightens and enriches the 
colors from yellow to red and a bluish illuminant like sky- 
light favors the violet, blue and green. The same argument 
applies to the complex colors. For example, purples reflect 
predominantly violet and red rays. Under a yellowish arti- 
ficial light they appear quite reddish as compared with their 
appearance under bluish light. It is interesting to try these 
simple experiments. One will learn much of the relation 
of quality of light to the appearance of colors by this means. 
That average noon sunlight is approximately white has 
been the general conclusion of those experienced in the 
science of color. However, I. G. Priest23 has recently con- 
firmed this by a thorough investigation. He found that 
white light would be produced by a true radiator if the 
temperature was approximately 8900 degrees F. He found 
average noon sunlight in Washington corresponded to a 
color temperature of 9080 degrees F. 
Ideal Illuminant for Color-Work 
From the foregoing discussion it should be obvious that, 
for color-work in general, the ideal illuminant is one which ILLUMINANTS AND COLOR 
99 
is quite neutral; that is, one which does not favor any 
colors. When we consider that, in a broad sense, we must 
distinguish between several thousand colors, and between a 
hundred thousand of them if their variation in shade is 
taken into account, the problem of color-discrimination is 
one that demands a proper illuminant. There is only one 
quality or spectral character of light that can qualify as a 
standard. This is a white light for it does not favor any 
colors. Noon sunlight on clear days does not vary appre- 
ciably in quality. It is approximately white. It is the most 
natural of our natural illuminants because of its overwhelm- 
ing power. It can be fairly well standardized over the 
world because it visits all the countries concerned with color- 
work. However, an artificial white light can be stand- 
ardized by proper color instruments. 
As already pointed out, when man came indoors the sky 
became the source of daylight upon which he depended. 
But as already explained, in the northern hemisphere, north 
skylight is less variable in quality and intensity than light 
from any other portion of the sky. Therefore, color-work 
gravitated to north windows but even here the light is vari- 
able and often insufficient for accurate color-work. Thus 
the idea that north skylight was the best light for accurate 
color-discrimination became firmly entrenched. It is so 
firmly rooted that it will be difficult to dislodge. There 
will be cases where it is the best quality, but they will be 
those where violets, blues, etc. are important and where 
there is an advantage in favoring them. But where true 
appearance of one color in relation to others is desired, 
white light will be standardized eventually. This is now 
possible because artificial white light is available. For ac- 
curate color-work the color-expert should adopt artificial 
white light and use it exclusively. Its cost is insignificant 
compared with its importance in saving time, in reducing 
spoilage and in guaranteeing satisfactory product. 100 
LIGHT AND WORK 
Accurate Artificial Daylight 
In the appearance of colors, the distribution of light and 
the color of the surroundings 19 are important. There are 
surface effects so that light from a brilliant source of small 
size will cause an appearance different from light from a 
large source. In general, the light should be diffused some- 
what and should come from a source subtending an appre- 
ciable angle at the object. This can be achieved by several 
artificial-daylight units in a special room with white ceiling 
and medium gray walls. 
A color is always influenced by its environment so that 
the latter should be the best compromise possible. This is 
a medium gray. A colored area seen amid black surround- 
ings appears relatively brighter than it actually is. Seen 
against white it appears darker than it is. There will be 
the least change in apparent brightness of the color under 
observation when the surroundings are equally as bright 
as the color. Therefore, in general, a medium gray is the 
best average or compromise. Besides this brightness-con- 
trast we have color-contrast. If the surroundings are of a 
certain color they generally tend to induce in the color 
under observation, a slight color approximately comple- 
mentary to their color. This is well demonstrated by 
placing a white or gray area in the midst of a colored field. 
If the color of this surrounding field is green the white or 
gray area will be seen to be tinged with a purplish color. 
Furthermore, we must contend with after-images. If we 
look intently at a bright color for a few moments and then 
look away at a white or gray surface, we shall generally see 
an area, the size of the bright color, tinged with a comple- 
mentary hue. If we look at a blue sky for a few moments 
and then at some colors, they will appear different than they 
do a little later. The effect of after-images in ordinary 
color-work is of relatively short duration - a matter of 
seconds - but when comparing two or more bright colors ILLUMINANTS AND COLOR 
101 
by looking from one to the other, the effect is noticeable. 
These aspects of light, color, and vision have been discussed 
in detail elsewhere.1 
If only a small spot is to be illuminated, a single artifi- 
cial-daylight unit is satisfactory. However, for the best 
discrimination of color, the intensity of illumination should 
be much higher than is usually found under ordinary arti- 
ficial lighting conditions. It should be at least 50 foot- 
candles if possible. In fact, here is a phase of vision where 
we should obtain the highest intensities practicable, for the 
eyes should be working under conditions as normal as pos- 
sible. Although the quality of skylight varies owing to 
changing sky-conditions, much of the variation of skylight 
entering a window is due to the changing lighting of the 
adjacent buildings due to variation of direct sunlight. 
Furthermore, colored surfaces such as brick walls, at one 
moment in direct sunlight and at the next in shadow, cause 
much uncertainty in fine color-work. Satisfactory artifi- 
cial daylight can be obtained over an area for one worker 
by means of a few hundred watts. Thousands of these 
lighting-units are now serving in a great variety of indus- 
tries involving accurate color-discrimination.28 Many 
operators and expert colorists are now independent of day- 
light which is very uncertain for color-work demanding 
accuracy. Accurate artificial daylight is now in use in 
textile mills, stores, cigar-factories, color-factories, paint- 
shops, garment-factories, cotton-exchanges, fur-industries, 
color-printing, art-studios, metal-work, refineries, chemical 
analysis, laundries, paper mills, flour-mills, sugar-refineries, 
jewel-shops, dentistry, microscopy, medical examination, 
surgery, woodwork, and in a great many other places. Any 
industry or human activity calling for the discrimination of 
color has uses for artificial daylight. Only a few of the 
different types of artificial-daylight units which have been 
placed on the market have been scientifically correct. The 102 
LIGHT AND WORK 
others have been makeshifts, or impracticable, or far from 
correct. These naturally eventually disappear from the 
market, but before they do many of them have found their 
way into the industries. An accurate artificial daylight is 
invaluable in many human activities; a spurious one is 
worse than none. 
Approximate Artificial Daylight 
The luminous efficiency of our practicable light-sources 
of the present time is not sufficient to produce accurate 
artificial sunlight for general lighting at a reasonable cost. 
However, in a machine-shop all the machinists do not need 
the finest set of calipers. Their work and their ability do 
not demand them. So it is with accurate artificial daylight. 
All workers do not need such an illuminant, but most of 
them would find advantage in an illuminant which approxi- 
mated noon sunlight or average daylight in quality. For 
these reasons the tungsten daylight lamp was made avail- 
able. Its light is a long step toward noon sunlight and 
everything appears more natural in proportion to this step 
toward noon sunlight. Millions of these lamps are in use 
and there is adequate proof of their desirability. 
Few persons who have not studied the subject, realize 
how much we unconsciously depend upon color in our every- 
day work. If we carefully study our surroundings or our 
work we will recognize that color plays an important part. 
Woods, metals, fabrics, papers and almost everything pos- 
sess color. In fact, it is difficult to find something that is 
colorless. For example, we may think a " white " powder 
is colorless but stick a finger into it and withdraw it. The 
hole will generally appear decidedly colored. The light 
entering the hole is trapped and reflected many times. Each 
reflection colors it slightly until it assumes a conspicuous 
tint by multiple reflections. A gold-lined goblet appears of 
a reddish color for the same reason. ILLUMINANTS AND COLOR 
103 
Many examples of costly spoilage are found which would 
have been eliminated even with the tungsten daylight lamp 
which is designed to give a light as near to average daylight 
for general lighting as cost of light will permit. It has been 
in existence for ten years and has invaded many fields of hu- 
man activity. It is useful not only where color-discrimination 
is important but also where daylight must be reinforced by 
artificial light. Here the reduction in color-difference be- 
tween the two illuminants is satisfying. The dominant 
directions of natural light and of artificial light are generally 
quite different. In such cases a marked difference in the 
color of the lights is very noticeable and quite annoying. 
This annoyance is much less when the two lights have the 
same general dominant direction as is the case of the arti- 
ficial-lighting luminaires being hung under a ceiling sky- 
light through which the daylight enters. Inasmuch as most 
daylighting is done by windows in the walls and artificial 
lighting by luminaires overhead, this unsatisfactoriness due 
to the two lights being different in color is quite common. Chapter VI 
QUALITY OF LIGHT AND THE HUMAN BEING 
THERE is a growing conviction in many quarters that 
the ideal illuminant from the viewpoint of the effi- 
ciency of the human being in everyday work, is white light 
or one approximately white. If we are to give due weight 
to the influence of adaptation to environment, certainly we 
cannot avoid the conclusion that light of a daylight quality 
- whether natural or artificial - is the ideal toward which 
to strive. Fortunately this is the tendency in light-produc- 
tion by the method of heating solid bodies, because as the 
efficiency increases a natural step toward white light is 
achieved. In practically all tests of the relation between 
the quality of light and the desirability of the light, it is 
found that the whiter light is chosen for work-purposes. 
Naturally many would like to know why this is true. To 
many of these the explanation involving evolution of the 
sense-organ and of the attitude of the human being toward 
light is not convincing. Perhaps, here and there, we may 
find a more material or a physical or an optical reason, but 
in the main the author believes that the most powerful and 
reasonable explanation is that of adaptation. This should 
be convincing to those who have an adequate conception of 
the invulnerable position that this principle has attained in 
the biological and psychological realms. We have not at- 
tained breadth of view in regard to man and his relation 
to the world in which he lives, until we recognize the great 
importance of physiological and particularly of psycho- 
logical phases of the human being. 
Nature's lighting has influenced us physically by having 
to do with the development of our physiognomy and of the 
104 QUALITY OF LIGHT AND THE HUMAN BEING 
105 
physical properties of the eye-media. It would take count- 
less centuries of slow adaptation to other lighting conditions 
to alter these physical characteristics so that they would be 
more suitable to the new regime of lighting. The physio- 
logical processes of vision are also a product of adaptation 
and doubtless would be altered very slowly to the new light- 
ing. Our mental attitude, or more broadly the psychological 
phase, is more susceptible to change. On every hand we see 
changes taking place within the short period of our lives. 
Furthermore, it is in this psychological realm where man 
differs from other animal life very much more than in the 
other more permanent or more enduring realms. We must 
take this into account in any broad and deep consideration 
of lighting. In doing so we iron out apparent inconsistencies 
and this is what any successful explanation must do. Our 
imagination cannot picture primitive man beginning his 
journey up the ladder of civilization without fire. Taking 
into consideration his use of fire and of its light and the 
conditions of its use, it is easy to conclude that this illumi- 
nant became natural to him under certain conditions which 
are responsible for certain associations and attributes. This 
phase of natural illuminants has been touched upon in 
Chapters IV and V and is further analyzed in the present 
chapter. 
Light of Natural Quality 
Where the discrimination of color is involved there is no 
doubt of the superiority of light of a white quality. This is 
subject to demonstration just as it can be demonstrated 
that it is better to eliminate color-difference when daylight 
and artificial lights have different dominant directions. 
However, further questions arise. What quality of light is 
best where color-discrimination is not involved? What 
quality of light are we inherently best adapted for? These 
are questions which will be completely answered eventually 106 
LIGHT AND WORK 
by research. We now have some data, a great deal of ex- 
perience, and certain logical philosophy. In viewing this 
aspect we should think of an average normal person as two 
beings psychologically. He is a day and a night being. 
With day and daylight, work and serious vision are asso- 
ciated. During the day we expect adequate intensities of 
light of daylight quality. This expectation is really an 
unconscious demand inherited perhaps from each ancestor 
of the long line of them extending far back of the veil of 
unrecorded history. This expectation or unconscious de- 
mand is fostered from the moment we begin to record 
experiences in early infancy. It becomes a most natural 
part of our lives and activities. 
Day with its high intensities of light of daylight quality 
is the dominant period of serious work. Our esthetic self is 
submerged by our working self. Night is quite the reverse 
for the average normal person. This is a period of rest and 
recreation in which the esthetic self may come forth. We 
may sit in the comfort and cheerfulness and restfulness of 
home or we may actively recreate. But in all this we have 
the spirit of home with which has been associated for count- 
less centuries the yellow-orange light from flames. This 
color of light has been indelibly stamped upon us as " night " 
beings. If our visual sense-organs are normal or at least 
not radically defective, our night self predominates and 
demands the color of light of home - that of flames. In 
the case of weak or peculiarly sensitive eyes, there is a 
natural demand for light of daylight quality, for this was 
the great overpowering environmental light to which the 
sense-organ became best adapted. Then for the present we 
will assume that for a perfectly healthy, normal visual 
sense, the psychological aspect of artificial light of the warm 
tint of flames dominates in our night environments indoors, 
at least when we are not engaged in work so serious that QUALITY OF LIGHT AND THE HUMAN BEING 
107 
our esthetic self is completely submerged. However, for 
the moment let us consider our " working " or " day " self 
in which the serious business of doing our part in the world's 
work is dominant. 
The meager data which we have on eye-fatigue and vision, 
in general, indicates that light of a daylight quality is more 
favorable to prolonged visual activity than light of other 
quality. In fact, there is an indication that light becomes 
generally more and more favorable as the quality approaches 
white. For example, for serious visual work in which the 
esthetic sense plays little or no part, as we increase the 
temperature of a tungsten filament the quality of light im- 
proves. Of course, all other conditions are assumed to re- 
main constant. Such a result is to be expected unless we 
are willing to admit that, only in the case of daylight as an 
environmental factor in evolution, the well established 
principle of adaptation fails. This the author cannot do; 
however, it is well to ascertain, if possible, whether or not 
actual data support the logic of the principle of adaptation. 
For one entire winter the author experimented with 
lights of different quality for serious work. At one end of 
the series was approximate daylight from the tungsten day- 
light lamp and at the other was a deep yellow-orange light. 
Between these were several steps of yellowish lights of 
the natural qualities of incandescent solids of different tem- 
peratures. In order to eliminate the " night " self or the 
dominance of the esthetic self, the experiments were con- 
ducted only while the author and another adult were en- 
gaged in the concentration of writing and computing 
attendant scientific and technical books and articles. Under 
such conditions one becomes quite oblivious of surroundings 
and of the spirit of comfort and cheerfulness of home. In 
other words, such concentration tends to keep the " work- 
ing " self dominant. Without any question the results 108 
LIGHT AND WORK 
pointed with favor to that illuminant which approached 
nearest to daylight in quality. The desirability of the dif- 
ferent illuminants decreased with increase of yellowishness. 
A more recent experiment performed with about 35 in- 
telligent persons seemed to be conclusive for the conditions 
of the experiment at least. The observers were persons 
engaged chiefly in clerical or laboratory work and none 
had any special interest in the question involved in the in- 
vestigation. Without going into detail it may be stated 
that the research was conducted in the approved manner 
of psycho-physiological researches in which extraneous in- 
fluences, suggestion, and prejudice were guarded against. 
The visual process chosen was that of reading black print 
on white paper such as the Saturday Evening Post. The 
surroundings were plain, ordinary, and practically neutral 
in color and played no important part in the results. The 
illuminants used were Mazda C (gas-filled tungsten) lamps 
with colorless bulbs and Mazda " daylight " lamps operat- 
ing at their normal voltages. The light in both cases was 
equally diffused and distributed to surroundings and reading 
matter in the same manner. Thus, there was only a dif- 
ference in quality with the single exception of intensity of 
illumination. The " daylight " lamps emit only about two- 
thirds as much light as the clear-bulb lamp of equal wat- 
tage. In order to make the investigation of the most 
practical value equal wattages of the two kinds of lamps 
were used. This made the intensity of illumination from 
the " daylight " lamps always only two-thirds of that from 
the regular clear-bulb lamps. This is not only a practicable 
aspect from the viewpoint of cost of light, but it was thought 
that if there was a tendency for superior quality to over- 
come inferior illumination intensity the results would be 
convincing. 
After a preliminary period of adaptation of the subject 
to the conditions, he (or she) was requested to read under QUALITY OF LIGHT AND THE HUMAN BEING 
109 
a certain level of illumination from one of the illuminants. 
A change from one illuminant to the other could be made 
at the will of the observer as desired in order to obtain a 
comparison or judgment. It will be noted that in reading 
the black print on white paper amid neutral surroundings 
the problem of discrimination of color did not enter at all. 
Here was a comparison of desirability of quality, pure and 
simple, although the factors involved in the chain of events 
from the stimulus to the retina to the brain and finally to 
consciousness need not be analyzed here. Notwithstanding 
the handicap of a lower intensity of illumination, the 
quality of light emitted by the Mazda daylight lamps was 
preferred by a majority of the observers as shown in Table 
VIII. 
TABLE VIII 
Results of an Investigation of the Preference of a Group of Observers for 
Two Illuminants, namely that from Regular Mazda C lamps and that from Mazda 
Daylight Lamps. 
Experiment 
Foot-candles 
No. of 
Observers 
Percentage 
Preferring 
Mazda daylight 
Mazda C 
Mazda daylight 
i 
0-75 
o-5 
28 
75 
2 
6,o 
4 
34 
62 
3 
135 
9 
32 
54 
4 
21.0 
14 
32 
54 
5 
28.5 
19 
34 
77 
Five levels of illumination were used and the observers 
were the same throughout. The research extended over 
most of a year and the matter of memory was easily elimi- 
nated. The levels of illumination extended from 28.5 foot- 
candles to as low as 0.75 foot-candles for the Mazda C 
(gas-filled) lamp and respectively two-thirds of these values 
for the Mazda daylight lamp. It is again emphasized that 
the latter gives only approximate daylight as explained in 
a preceding section. 110 
LIGHT AND WORK 
The most significant result is that the majority in all 
cases preferred the light from the approximate daylight 
quality over that from the regular gas-filled tungsten lamp 
even with the handicap of equal wattage. No great weight 
can be given to the variation of the percentages in the last 
column but it is extremely interesting to find that even a 
low intensity as 1/2 to 3/4 foot-candles, where light might 
seem to be of more importance than its quality, the Mazda 
daylight lamp was favored by three out of four persons. 
Another experiment was carried out whose results even 
more emphatically favored the light of approximate day- 
light quality. It has been commonly observed that mixing 
ordinary artificial light with natural daylight, differing both 
in direction and quality, is unsatisfactory, particularly when 
the quantity of one was of the same order of magnitude 
as the other. The same set of intensities was used, from 
the same two overhead sources (Madza C and Mazda day- 
light lamps of equal wattage), but at all times 4 foot-candles 
of natural daylight from the sky through a window at the 
left illuminated the reading matter. The results are pre- 
sented in Table IX. 
TABLE IX 
Showing the Preference of a Mixture of Approximate Artificial Daylight with 
Natural Daylight over a Mixture of Light from Equal Wattages of Tungsten 
Gas-filled Lamps with the Natural Daylight. 
Experiment 
Foot-candles Added to 4 Foot- 
candles of Natural Daylight 
No. of 
Observers 
Percentage 
Preferring 
Mazda daylight 
Mazda C 
Mazda daylight 
6 
6.0 
4 
32 
75 
7 
13-5 
9 
36 
67 
8 
21.0 
14 
36 
72 
9 
28.5 
19 
36 
89 
These results show that the light of approximate day- 
light quality from Mazda daylight lamps is more favored QUALITY OF LIGHT AND THE HUMAN BEING 
than even the greater amount of light obtained from Mazda 
C lamps of equal wattage. From previous work it has been 
proved that, within the range of intensities considered here, 
persons engaged in reading favor more light in each case 
if it is available. Therefore, it appears to be well estab- 
lished that, for equal intensities of light from Mazda C and 
Mazda daylight lamps, the latter light which approximates 
daylight in quality meets with decidedly more favor than 
the former for close visual work, assuming, of course, that 
general conditions are approximately the same. 
The results of long experience are valuable to the indi- 
vidual and through them he is guided in practice. Data 
on the influence of quality of light on us as human beings 
are very difficult to obtain. Not only is this psycho-physio- 
logical realm a trying one for the investigator, but in most 
cases there are overwhelming difficulties in obtaining a suffi- 
ciently large group of subjects who will devote themselves 
to the ordeals. Furthermore, in this particular field the 
results obtained from only one or two observers are often 
unconvincing, owing to the great variation among human 
beings in their reactions to light. However, the weight of 
evidence is bound to accumulate with experience and such 
is the case with the author during many years in this field. 
These indefinite data will not be convincing to many so 
that they will not be discussed in detail. 
Throughout the past ten years or more, many cases of 
individuals with over-sensitive eyes have come to the 
author's attention. Usually the statement is made that 
artificial light " hurts " their eyes. Knowing how careless 
most persons are with artificial light, one is always justified 
in assuming that there is improper distribution, lack of 
shading or diffusion, or insufficient intensity. However, 
after eliminating these factors we have plenty of evidence 
that many persons cannot read with comfort under ordinary 
artificial light at night who read with comfort under arti- 
111 112 
LIGHT AND WORK 
ficial daylight. In most of these cases the approximate 
daylight from the tungsten daylight lamp is satisfactory. 
The testimony in this direction is plentiful and authentic. 
This leads us naturally to a seeming contradiction but 
which disappears on analysis. Many persons express dis- 
like for the " white " light from ordinary tungsten lamps 
and speak in glowing terms of the warm yellow tint of flames 
or firelight. In the first place, they only think the former 
light is white for it is quite yellowish compared with sun- 
light or skylight. The same persons make no complaint 
of the whiteness of daylight during their period of work 
or serious activity. This inconsistency rules out of con- 
sideration the complaint against the " whiteness " of 
ordinary artificial light unless we recognize that man is 
psychologically a different being by day than he is at night. 
Evidence on every hand strengthens this conclusion. 
Usually this complaint of " whiteness " of ordinary artificial 
light arises in the home or in environments having the home 
atmosphere. The period of work has passed and the esthetic 
self comes forward as already suggested. Now the human 
being rests, or wants to, even while engaged in definite 
activities such as reading, sewing, etc. Even in these pur- 
suits the factor of recreation is prominent. After our work- 
day, our interiors are havens of comfort, recreation, and 
cheerfulness. Firelight has ever been a symbol of such 
havens. We cannot easily overcome this instinct for light 
of warm tint. Of course, there are unsensitive beings to 
whom this is piffle, but their process of thought and their 
outlook on life which have made them unsensitive to the 
esthetic world have also made them so unhuman in so 
many ways that their opinions are of little moment in com- 
parison with those of human beings. Persons who are not 
developing in a normal manner psychologically are at the 
rear of the parade of human progress. 
On the other hand, super-esthetic persons are interesting 
subjects. To them firelight is full of charm and ordinary QUALITY OF LIGHT AND THE HUMAN BEING 
113 
artificial light is garish and harmful. They declare em- 
phatically that the latter hurts their eyes terribly and that 
they cannot stand it at all. After due analysis and argu- 
ment one is generally able to conclude that the whiter lights 
hurt merely their esthetic self but this self is so dominant 
they think that the light is really physiologically harmful. 
These same persons actually enjoy full daylight in their 
summer garden without even a thought of its real white- 
ness. 
It is hoped that these few glimpses of the psycho-physio- 
logical effects of light may aid the reader in understanding 
and particularly in analyzing this complex aspect of light and 
lighting. The author firmly believes that his assumption of 
two different beings or attitudes for each normal person is in 
general correct. Certainly, it explains many observations 
and irons out some apparent inconsistencies. Long ex- 
perience and observation have established this conviction 
and as already stated much of the data is of that indefinite 
nature which cannot be transmitted to others without labor- 
ing far afield. 
Light of Unnatural Quality 
From the preceding discussions and Fig. 19, it is seen 
that the chief characteristic of a natural iliuminant is a 
continuous spectrum in which visible radiations of all wave- 
lengths within this spectral region are present. There is a 
great difference between the actual distributions of these 
radiations throughout the visible region for sunlight and 
candle-light. But it has been noted in Table VII that the 
various electric filament lamps take logical places in the 
series between candle-light and sunlight. These artificial 
illuminants are due to glowing solids of various tempera- 
tures and in a sense arrange themselves in a natural scale 
as to color. In fact, their color is not conspicuous because 
of this degree of naturalness. Doubtless, the reader will 
recall the artificial illuminants which seem unnatural. The 114 
LIGHT AND WORK 
gas-mantle after it has been in use for some time assumes 
a greenish tinge. This is noticeable and has attracted un- 
favorable comment. However, a redeeming feature of the 
light from a gas-mantle has been its continuous spectrum; 
that is, the presence of all wavelengths of visible radiation. 
It is greenish because of the presence of an abnormal 
amount of radiations of wavelengths in the green region of 
the spectrum. However, the light from the gas-mantle is 
only tinged with " unnaturalness," and having a continuous 
spectrum it passes muster in any group of illuminants. 
The only radically unnatural illuminant in appreciable 
use at the present time is the light from the mercury-arc. 
Its total light is bluish white; that is, a gray or colorless 
surface will appear bluish when illuminated by its light. 
However, its quality or spectral character as shown in Fig. 
19 consists of four principle lines M in the visible region 
and to two of which, the green and yellow (double), more 
than 90 per cent of the total light is due. Thus this light 
is almost as " unnatural " (in the sense of the term as used 
in this book) as an illuminant could be. The least natural 
would be one that had a single line in the extreme portions 
of the spectrum, say, red or blue. Incidentally this would be 
known as monochromatic light because the light is confined 
entirely to one spectral hue or line. Therefore, mercury- 
arc light approaches monochromatism to the extent that 
the total light approaches in spectral character a very nar- 
row band of a single wavelength. Such a light is bound to 
have peculiar properties and, therefore, a brief discussion 
of these is presented. 
Perhaps the most noticeable characteristic of mercury-arc 
light is the peculiar and unnatural appearance which most 
colored areas assume when illuminated by it. As discussed 
in a previous section, red objects appear practically black 
because this light has no orange and red rays in it. It is 
certain that we would not illuminate everything with a 
blue light or a green light made by coloring tungsten-lamp QUALITY OF LIGHT AND THE HUMAN BEING 
bulbs, for example. Fortunately, there are sufficient blue 
and violet rays in the mercury-arc light, which when mixed 
with the green and yellow rays give a gray or colorless object 
a bluish color not differing radically from white. This 
makes the light from the mercury-arc endurable even though 
it causes most colors to appear unnatural. 
Owing to a certain shortcoming of the eye (chromatic 
aberration), very fine details near the limit of visibility are 
seen under monochromatic light at a somewhat lower in- 
tensity of illumination than in the case of ordinary light. 
Dr. Louis Bell20 and later the author 1 studied this matter 
thoroughly. It was found that at the limit of visibility, the 
revealing power of mercury-arc light is superior to sunlight 
or any light of extended spectral character but that this 
advantage decreases as larger details are considered. The 
advantage rapidly decreases as the intensity of illumination 
increases. It practically disappears for ordinary visual ac- 
tivities such as reading under adequate illumination inten- 
sity. Very seldom do we call upon the eyes to distinguish 
details so small as to be near the limit of visibility. The 
engraver, for example, is engaged with finest detail more of 
the time than most other persons; still, engravers often pre- 
fer light of a daylight quality. In one of the largest printing 
establishments in this country where mercury-arcs are used in 
some parts of the printing plant, the engravers use natural 
and artificial skylight. The advantage of monochromatic 
light has been over-rated by ignoring the fact that we gen- 
erally are not required to distinguish very fine details near 
the limit of visibility for the particular lighting conditions. 
Of monochromatic lights it is fairly well established that, 
for the normal eye, those yellow and green regions are the 
best from the viewpoint of visual acuity. The mercury-arc 
is fortunate in this respect because most of its light is due 
to the green and yellow spectral lines. 
There are many special applications of mercury-arc light 
but it is questionable whether or not a group of workers 
115 116 
LIGHT AND WORK 
should be subjected to the unnaturalness of appearance of 
themselves and surroundings. In special cases a light of 
such peculiar properties can be utilized for revealing certain 
things. However, the advantage of such light can best be 
determined by experiment, although a knowledge of the 
science of color helps in making predictions.1 Some claims 
have been made to the effect that reaction time is less under 
mercury-arc light than it is under other illuminants; how- 
ever, the data are not convincing because of the insufficient 
number of observers and for other reasons. With great 
respect for the wonderful achievements represented by com- 
mercial mercury-arcs, the author believes their place in the 
world's work is in very special applications and not for 
general lighting where large numbers of persons must be 
subjected to the psychological influence of unnatural 
appearance of people and of their surroundings. Certainly 
this question should be settled if possible by very careful 
research involving many workers. 
We are subject to influence by our environment. No 
person even though he has the ability to observe, to think 
and to remember can wholly resist his environment. This 
is contrary to the principles of adaptation. It seems rea- 
sonable that a person cannot remain unaffected by the 
ghastly appearance of his co-workers under mercury-arc 
light. It seems reasonable to expect that depression or 
distortion of some kind might arise from the general un- 
naturalness. It seems reasonable that one's idea or ap- 
preciation of color would be distorted after hours each day 
spent in such an environment. All these factors can be stud- 
ied but the amount of work involved would be very great. 
One is not encouraged to do this when he is fortified with illu- 
minants which are in the natural scale and with the logic 
that mercury-arc light is too radically different from natural 
illuminants to assume the responsibility of using it, when 
there are no prominent advantages of cost or of general 
visibility. QUALITY OF LIGHT AND THE HUMAN BEING 
There had been so much discussion of the advantages and 
disadvantages of a light approaching monochromatism, that 
a simple experiment was devised to see how much light a 
reader would choose of two radically different qualities.22 
A box with a bottom of diffusing glass and containing gas- 
filled tungsten lamps was hung overhead by the side of 
a mercury-arc. Brightnesses of visible portions of light- 
sources were of the same value and the distribution and 
dominant direction of light were practically identical for 
the two illuminants and the surroundings were neutral. 
In other words, all conditions were identical excepting 
quality or spectral character of light. The reading material 
was in all cases black print on white paper. After an adap- 
tation period of reading at a selected intensity of illumi- 
nation, the reader was permitted to change the intensity 
of illumination as desired. For each experiment with the 
two illuminants a certain maximum intensity was obtain- 
able. These maximum intensities were io, 30, and 45 foot- 
candles. The condensed data of the investigation are pre- 
sented in Table X. 
117 
TABLE X 
Choice of Intensities of Illumination of Lights from Mercury-arc and Gas- 
filled Tungsten Lamps. 
Test 
No. 
Number of 
Observers 
Maximum 
Foot-candles 
Obtainable 
Average Foot-candles Chosen 
Mercury-arc 
Gas-filled 
Tungsten 
i 
22 
IO 
6-3 
5-3 
2 
55 
3° 
14. 2 
12.7 
3 
24 
45 
16.9 
16.1 
4* 
26 
30 
18.0 
17-4 
* In this case the paper was dyed gray of 41 per cent reflection-factor. 
It is seen that there is no marked difference in the inten- 
sity chosen. It seems likely that if one illuminant was very 118 
LIGHT AND WORK 
much superior to the other in promoting visibility, less of 
this light would generally be chosen in comparison with an 
illuminant of inferior quality. As the maximum available 
intensity increased the average illumination chosen also 
increased. Why the maximum intensity was not chosen 
may be explained in the following manner. Before be- 
ginning the investigation, which ran for a period of several 
months, it was expected that a subject would likely choose 
an illumination somewhere between the maximum and zero. 
Not knowing what was the underlying object he might 
think that the best condition was somewhere between the 
maximum and minimum available. However, there was no 
way to avoid this and besides the same attitude applied to 
both illuminants. The observer made his choice of illumi- 
nation for each of the illuminants for only one available 
maximum illumination at a setting. He was tested for the 
condition of another available maximum sometimes weeks 
later. Considering this method of procedure, the large num- 
ber of individuals, the several different " sittings " of each 
individual, the large total of observations and the fact that 
he could make his choice of illumination for each illuminant 
as he desired, it is surprising that there is no marked dif- 
ference in the average choice of illumination for the two 
illuminants when their maximum available intensities were 
the same. If there was a marked difference in the " reveal- 
ing power " of these two illuminants the results should have 
shown some indication of it. Besides showing no marked 
difference in the choice of illumination intensity, the re- 
sults indicate that there is no marked advantage of mercury- 
arc light in promoting visibility of details like reading 
matter under adequate illumination intensities, and that 
there is no low limit of desired illumination intensity. It 
will be noticed that slightly more mercury-arc light was 
chosen on the average than light from gas-filled tungsten 
lamps. Chapter VII 
QUANTITY OF LIGHT 
ILLUMINATION intensities have been discussed here 
and there in preceding chapters but they will be dis- 
cussed briefly here more specifically, even at the risk of some 
repetition. In the matter of intensity of light we have a 
curious anomaly - perhaps we should say inconsistency - 
in human activities. As has been pointed out, we have 
enormous intensities of illumination outdoors during the 
day, but our lighting at night is very close to the other ex- 
treme - darkness - by comparison. Still we often hear 
complaints of " overlighting " and " too much light," when 
the actual intensity of illumination is only a few foot- 
candles. The same intensity on our porch a half-hour after 
sunset seems feeble indeed and unless our book is par- 
ticularly fascinating we lay it down to await the artificial- 
lighting period indoors. When we go into the house some 
time later, we may find a condition which sometimes 
arouses the complaint of " too much light." Measurements 
would show the extreme folly of the remark, but there is 
another way of doing this. If the room is well lighted in 
the daytime, turn on the artificial lighting and its influence 
on the illumination intensity will scarcely be noticeable. 
If it cannot make its influence felt upon this well-daylighted 
interior it would be insignificant in competition with the 
daylight outdoors which is hundreds of times more intense 
than the daylighting indoors upon which it could scarcely 
make an impression. The truth is that glare from unshaded 
lamps and undiffused light is confused with quantity of 
light. It is also interesting to note that fine work generally 
119 120 
LIGHT AND WORK 
is located near windows in factories and elsewhere. In actual 
practice we cannot obtain too much light indoors if the 
lighting conditions are proper. Nearly every visual law 
shows that the efficiency of the eye, or at least its ability to 
see, increases as the intensity of illumination increases up to 
hundreds and even thousands of foot-candles. 
The Sun as a Light-Source 
In many respects the aim of human progress is either to 
improve upon or to become independent of Nature. Owing 
to the inadequacy of daylight indoors in many places in our 
congested cities, electric lighting is a great improvement over 
natural lighting, but when we think of the sun as a light- 
source, and of the intensity of illumination which it pro- 
vides over one-half of the earth's surface constantly, the 
accomplishments of mankind in lighting are feeble indeed. 
There is such a general misconception of the relative intensi- 
ties of natural and artificial light that the overwhelming 
intensity of sunlight compared with ordinary artificial light- 
ing intensities cannot be over-emphasized. As has been 
pointed out, the luminous intensity of the sun is over two 
billion billion billion candle-power as measured at the 
earth. All the artificial light ever made by mankind 
dwindles to an insignificant spark by comparison. No ade- 
quate conception of this enormous luminous intensity can 
be obtained from the bare figures; however, we may put it 
in terms of the earth's population of about two billion 
beings. If we gathered the earth's population into one 
group and gave each person a billion lamps of a billion 
candlepower each we would have an aggregate light-source 
which would compete with the sun in luminous intensity. 
Even now we cannot conceive the magnitudes, so let us 
try again. If the present rate of manufacture of artificial 
light-sources were increased a billion times it would still 
require 100 million years to accumulate enough lamps for QUANTITY OF LIGHT 
their combined candle-power to compete with that of the 
sun. Such figures may shock us into a realization that there 
is no danger of over-lighting this earth with the feeble 
facilities of mankind at the present time. 
Now let us approach the matter differently by eliminat- 
ing the great distance of the sun and consider illuminating 
the surface of the earth from a reasonable earthly distance 
by means of lamps contained in perfect reflectors. Sup- 
pose we distribute the population of the earth uniformly 
over its surface. Each person would alone occupy a 1700- 
foot square. If present-day, 1000-watt tungsten lamps 
were used and all the light were reflected toward the earth, 
each 1700-foot square would require a million 1000-watt 
lamps to illuminate it to an average daylight intensity dur- 
ing midday. In other words, each person on the average 
would have to manufacture one million 1000-watt lamps 
every 1000 hours, or 25,000 lamps each day. 
121 
Artificial Lighting and Decreasing Cost 
Such figures are staggering but they show what a job 
mankind has to accomplish before he is going to have the 
earth overlighted. Of course, the comparison is unfair 
because outside of streets and a few other places mankind 
is interested only in lighting indoors. When we consider 
our indoor world we have an area considerably shrunken 
as compared with the earth's surface. The lighting expert 
has been conservative and has had his eye on costs. Per- 
haps his attitude, even though it has been a fair one, has 
not been best for mankind. His recommendations for gen- 
eral lighting indoors average only a few foot-candles. He 
realizes that adequate artificial lighting became possible 
only recently and that he must contend with the inertia of 
the people who not long ago accepted the light from candles 
and have not broken the habits of those relatively dark 
ages. In one-half century light-sources have developed 122 
LIGHT AND WORK 
from feeble flames to sources hundreds of times more 
powerful. 
The cost of light has decreased so enormously that we 
can purchase today as much light for one cent as we could 
a century ago for one dollar. We have a flexible dollar 
which is often in a shrinking mood and we have a steadily 
increasing standard of living. It is true that we are using 
much more artificial light than we did a century ago, but 
this increase is not in proportion to the decrease in cost 
of light. When we consider the tremendous increase in 
indoor activities - the great commercial and industrial 
activities - we can come to no other conclusion than that 
mankind has not taken advantage of decreased lighting cost 
in proper proportion. Certainly the workman of a century 
ago had an intensity of illumination of a foot-candle if he 
was obliged to do fine work after dark. Would the same 
workman for the same work, have 50 to 100 foot-candles 
today even though he could have it at the same cost? The 
answer is that relatively few workmen are provided with in- 
tensities of illumination of this magnitude, although it is 
very likely such intensities will be common-place in the near 
future. 
Perhaps at this point it should be stated that, with a 
properly designed lighting system and proper surroundings, 
there is no intensity of illumination, obtainable today in a 
practicable manner, that would be too great. Most of the 
visual processes become more and more efficient as the in- 
tensity of illumination increases up to hundreds of foot- 
candles. Under proper conditions, such as exist outdoors 
in nature, we shall not have our work-places overlighted 
even at an illumination intensity of a thousand foot-candles. 
It is shown in Chapter VIII that our ability to see fine 
detail and brightness-difference steadily increases up to very 
high brightnesses and illumination intensities. Let us at- 
tempt to compare the present state of affairs with that of a QUANTITY OF LIGHT 
123 
century ago. Let us estimate that the purchasing power 
of our money is only one third as great as a century ago. 
Let us guess that our standard of living and working has 
increased in general five times during the past 100 years. 
The last item is a guess because it can scarcely be appraised. 
Of course, anyone is at liberty to revise these two guesses; 
however, on this basis we should have about 1500 times 
more artificial light for our work-places than would have 
been available for workmen 100 years ago. 
It requires no unusual imagination or ability to observe, 
to conclude that artificial lighting has not kept pace with 
progress as represented by increase in efficiency of light- 
production and in the standard of living and working. Only 
by realizing the benefits of adequate and proper lighting 
will artificial lighting eventually come fully into its own. 
The lighting expert does not demand at the present time in- 
tensities of illumination of a thousand foot-candles. He 
would be happy to see an average of 20 or 30 foot-candles 
for work-places at the present time and would be confident 
that such a standard would convince the manufacturer and 
others of the great value of adequate and proper lighting. 
When he is convinced, higher standards will eventually 
come more or less as a matter of course. 
Natural and Artificial Light Indoors 
It has been pointed out several times that mankind by 
coming indoors placed many handicaps on natural lighting. 
Excepting from overhead skylights of proper design, the 
distribution of natural light indoors is not very satisfactory. 
The non-uniformity in large side-lighted rooms is very great 
and the horizontality of the dominant direction of natural 
light at some distance from windows is not satisfactory ex- 
cepting in very special cases.24 The intensities of illumina- 
tion are also very much less than outdoors. In fact, they 
are so low that it is easy to attain them by means of arti- 124 
LIGHT AND WORK 
ficial light. Natural lighting indoors costs about as much 
as artificial lighting. Considering the advantages of con- 
trol and constancy of artificial light, it is seen that it is a 
formidable competitor of natural light indoors. 
By no means are daylight intensities as great in factories 
as many are likely to believe from their experience at home 
or in well-lighted offices. Ward Harrison 25 made a survey 
of lighting conditions on clear and cloudy days in repre- 
sentative factories of various industries. From his sum- 
mary it is seen that the average illumination of horizontal 
surfaces was less than io foot-candles. In fact, in only 
two out of seventeen kinds of industries visited was the 
average above ten foot-candles. It varied generally from 
five to ten foot-candles on a horizontal surface at the work- 
places. The diversity of the illumination on vertical sur- 
faces was greater, being from two to thirty-five foot-candles, 
with an average somewhere in the neighborhood of ten 
foot-candles. These are intensities of illumination obtain- 
able at reasonable cost by means of artificial light and as is 
shown in Chapters X and XIV with a considerable increase 
in production and other desirable factors quite in excess of 
the cost of artificial light. 
Artificial lighting is necessary for economical operation 
of a factory because it must be available when daylight 
fails, as it is likely to do at any time. This should be 
charged to natural lighting. But we also have the possibility 
of using the same artificial-lighting installation for night- 
work. With the increasing standard of living we may be 
driven to two-shift operation in order to put brakes on the 
mounting cost of living. It is true that night-work has not 
been considered favorably in general. A part of this dis- 
favor is due to experience with inadequate artificial light. 
However, work-conditions can now be fully as satisfactory 
at night as during the day and perhaps the two-shift plan 
of eight hours each will be worked out eventually. Such a QUANTITY OF LIGHT 
125 
plan would distribute overhead over twice the output with 
an obvious advantage to all concerned. Mr. Harrison25 
computed the cost of artificial lighting to equal the average 
daylighting intensities and his conclusions are presented 
herewith. 
" To produce an illumination varying from 5 to 10 foot-candles in 
the different departments of a factory would under present conditions 
require an expenditure of a little more than 1 watt per square foot 
of floor space, assuming that well-diffused lighting units were to be 
employed. With current at 2 cents per kw. hr. the total cost of lighting 
should not exceed 7.5 cents per square foot per year for a night 
shift of 2,500 hours. 
"If, on the other hand, doubling the size of the factory and oper- 
ating on a day shift only were to be decided upon we would have to 
compare with the 7.5 cents per square foot annual cost of lighting 
the annual rental value of the addition to the plant fully equipped 
with machinery and ready to operate. Typical data on this point 
show a range for various plants of 30 to 75 cents per square foot 
per year, annual charge to cover interest, depreciation, and taxes on 
the plant and its equipment. The rental value depends somewhat 
on the location and type of construction but largely upon amount of 
machinery required. Thus the charge for a plant of 100,000 square 
feet area would fall between $30,000 and $75,000 depending upon the 
industry. The cost of lighting this plant for night shift operation would 
not exceed $7,500. This saving $20,ooo-$7o,ooo represents the dif- 
ference between success and failure for many companies. To look at 
it from another viewpoint, the consumer's viewpoint, rent now com- 
prises from 15 to 30 per cent of the increase which accrues to raw 
material in passing through a manufacturing establishment. Night 
operation should effect a substantial reduction in the cost of manu- 
factured articles. But granting these facts, there remain the objections 
often raised that men do not like to work at night, that good men do 
not have to work at night, that it is impossible to change the estab- 
lished mode of life, and that really restful sleep cannot be secured 
during the day. Of the four, this last objection is a valid one; in 
fact, it is an open question whether satisfactory service can ever be 
obtained from a shift which begins work at midnight. On the other 
hand, man's preference for working during daylight is in many cases 
nothing more than an outgrowth of the fact that until recently effec- 
tive work could not be done under artificial lighting, and that amuse- 126 
LIGHT AND WORK 
ments have gradually come to appropriate that part of the day which 
was otherwise unoccupied. If factories were operated on a morn- 
ing and evening shift - the latter ending by midnight - the ordinary 
hours of rest would be but slightly disturbed, and amusements 
which are dependent for their success upon a man's leisure time and 
his money would soon adapt themselves to the change. Under these 
conditions one would anticipate much less objection to night em- 
ployment." 
Influence of Daylighting on Artificial Lighting 
The most striking evidence of the limitations which have 
been imposed upon natural light by the indoor life of man 
is the relatively high intensities of illumination outdoors 
at times when natural light indoors is considered inadequate. 
In other words, from the viewpoint of our indoor activities, 
darkness comes much before it does outdoors; that is, 
" day " indoors is much shorter than it is outdoors. Mr. 
A. Smirnoff26 made an investigation of this in Washington, 
D.C. In this city manufacturing is a small part of the 
central-station load. Owing to the fact that the business- 
district supply of electric energy could be recorded inde- 
pendently of the residential supply, he applied his observa- 
tions chiefly to the former. He obtained continuous records 
of daylight throughout various days and compared these 
with the current readings in the central-station. His results 
apply to the business district as a whole and therefore are 
interesting as mean values in this particular locality. Mr. 
Smirnoff by observing the current recorder could observe 
the variation of artificial lighting with variation in daylight. 
He found that there was a definite increase in consumption 
of electrical energy when the records of the Weather Bureau 
in the city showed that total daylight, at a point where the 
entire sky was visible, decreased below 1500 foot-candles. 
As the intensity of daylight illumination outdoors in an 
unobstructed location decreased below this value the cur- 
rent reading increased, but all variations in outdoor day- QUANTITY OF LIGHT 
127 
light illumination above 1800 foot-candles had only a 
negligible influence. Some of Mr. Smirnoff's data are 
reproduced in Fig. 20 where it is interesting to note the gen- 
eral regularity of the current-consumption curve on a clear 
day (July 25) with that for July 29 when it became quite 
cloudy in the afternoon. The dotted record below these 
two ampere-records represents the variation of total day- 
light illumination on a horizontal surface in the open on 
6000 
5000 
% 
Increase in 
electric lighting due 
to cloudiness 
AMPERE5 FOR LIGHTING 
4000 
c\»°r 
9600 
8400 
7200 
6000 
4800 
3600 
2400 
1200 
FOOT-CANDLES OUTDOORS 
3000 
zooo 
1000 
0 
0 
8 9 10 II NOON 12 3 4 5 
TIME OF DAY 
Fig. 20. Showing the influence of cloudiness on the amount of electric 
current used for artificial lighting. The dotted line represents the daylight 
illumination intensity outdoors (right-hand scale) on a horizontal surface 
exposed to the direct sunlight and the entire sky on July 29. The two other 
curves are total current consumption for a cloudy day (July 29) and a clear 
day (July 25) respectively. 
July 29. The horizontal broken line at 1500 foot-candles 
indicates the level of outdoor illumination at which day- 
light becomes inadequate in some portions of the indoors 
which have been depending on natural light. 
It will be noted that the difference between the two 
records of current consumption on the clear and cloudy 
days respectively is considerable. It represents expendi- 
ture for artificial light during the period that daylight is 
expected. It is a charge that might legitimately be placed 128 
LIGHT AND WORK 
against daylight. This amounts to a great sum in our con- 
gested cities. In Fig. 20 it is seen that the consumption of 
electric energy due to the cloudy afternoon of July 29 was 
about 20 per cent greater than on a clear day. When costs 
of artificial lighting are computed, that due to failure of 
daylight in the daytime should be subtracted and should 
be added to the cost of daylight. Of course, the illumination 
value outdoors at which daylight begins to fail indoors varies 
extremely for various interiors depending upon exposure, 
visible sky-area, etc. When the 1500 foot-candle point was 
reached in the open in Washington, Smirnoff found that the 
illumination at a certain west window was 360 foot-candles 
but only 9 foot-candles 6 feet away from it. In many 
other places it was much less. Assuming that artificial light 
is turned on when daylight illumination indoors reaches 
two foot-candles on the average, one arrives at the conclu- 
sion that the maximum average daylight illumination in- 
doors is about 10 foot-candles. This value coincides fairly 
well with general experience. Thus it is seen that average 
illumination intensities of daylight indoors in our congested 
cities are not high. In fact, they average less than one per 
cent of outdoor illumination intensities, being of the order 
of magnitude of one hundredth to one thousandth. They are 
of such low value that they can easily be attained by arti- 
ficial lighting at a comparable cost. 
Without considering cost, the ideal system of lighting 
designed wholly from the viewpoint of the greatest cer- 
tainty, speed, safety and comfort in human activity will 
provide an illumination intensity of hundreds of foot- 
candles (See Chapter XIII). Even considering the matter 
of present cost of lighting, the indication is that increased 
production, greater safety, less spoilage, and better esprit 
de corps should justify intensities of illumination many 
times greater than are now in general use (See Chapter 
XIV). QUANTITY OF LIGHT 
In a survey of industrial lighting made in 1919 it was found 
that about one-third of the total operation of industrial plants 
is under artificial lighting. This means that an enormous 
value of products of manufacture is produced under arti- 
ficial light. The survey indicated that in only 40 per cent 
of the factories visited was the lighting such as could be 
pronounced good as judged by standards of 1919. In- 
cidentally, the standard which represents adequate and 
proper lighting as established by lighting experts has con- 
tinually increased during recent years. In view of what is 
known of light and its relation to visual activity, it may be 
stated with complete safety that there is not a factory in 
the world today illuminated to the point of maximum 
productiveness. 
129 Chapter VIII 
FUNDAMENTALS OF VISION 
VISION is accomplished by our ability to distinguish 
differences in brightness and in color. Although color- 
vision plays a much more important part in every-day vision 
than most persons are conscious of, the real fundamental 
of vision is the discrimination of brightness-difference. If 
an object has the same brightness as that of its background 
and the brightness is uniform in both cases, the object can- 
not be seen. In rare instances this may be true but there 
may still be a difference in color. In such cases, color-vision 
makes it possible to distinguish the object. In the absence 
of color-vision mankind could go about every-day work 
quite unhampered, but, of course, some other provision 
would have to be made in those cases where color has been 
used for specific purposes such as in signals. The color- 
blind person does not experience the charm and the interest 
in the magical drapery of color which color-vision makes 
it possible to enjoy. By no means is it to be understood 
that the author relegates color to an unimportant place in 
every-day vision. It is important but not as fundamentally 
important as brightness. Therefore the discrimination of 
brightness, of brightness-difference, and of colorless details 
is emphasized more than color in these discussions and data. 
The subject of color is touched upon only when it plays a 
fundamental part but it is so complicated that it can not be 
treated as completely as it has been elsewhere.1 
The Eye 
Our ability to see fine details, in other words our visual 
acuity, depends primarily upon certain optical properties 
130 FUNDAMENTALS OF VISION 
of the refracting media of the eye, the structure of the ret- 
ina, the intensity of illumination, the relative and absolute 
brightnesses of the object and its background, and the size 
of the object or the visual angle subtended by it at its dis- 
tance from the eye. The spectral character of the illumi- 
nant, and the colors of object and background, play parts 
that have been discussed in Chapter V and elsewhere.1 Our 
ability to distinguish brightness-differences of relatively 
large areas is also influenced by the same factors as our 
ability to see fine details excepting that retinal structure 
131 
B 
c 
A 
D 
Vitreous 
Humor 
K 
/AQUEOUS • Hunotl 
Opfiail axis' 
L 
G 
F 
J 
I 
H 
Fig. 21. Illustrating the various parts of the eye by a vertical cross-section 
along the optical axis. 
A Conjunctiva 
B Retina 
C Choroid 
D Sclera 
E Fovea 
F Blind spot 
G Optic Nerve 
H Ciliary body 
I Iris 
J Cornea 
K Suspensory ligament 
L Lens 
and, in fact, the resolving power of the eye as a whole, are 
generally less important than in the case of distinguishing 
fine details. A diagram of the eye is shown in Fig. 21 
with the various parts indicated. 
The eye must focus a fairly accurate image on the retina 
if the object is to be seen clearly. In this respect the de- 
fects and limitations of the optical parts of the eye are those 132 
LIGHT AND WORK 
which are to be looked for in any more or less simple optical 
system. The retinal structure or the smallness of the in- 
dividual elements responding to light also influences the 
clearness of fine details and, in fact, limits the smallness 
of size of details which are distinguishable. Its limitations 
are similar to the " grain " of a photographic emulsion. 
When we inquire into the eye as an optical system we con- 
clude that it is far from perfect but when we think of its 
great range of sensibility, or its adaptability to a great range 
of brightnesses, we find it a very remarkable sense-organ. 
In fact, we agree with the great Helmholz when he said that 
he could make a better optical instrument but not a better 
eye. 
The sensibility of the eye depends upon such factors as 
the state of adaptation and the visual angle of the stimulus. 
As the visual angle increases, there is a decrease in the 
minimum amount of radiant energy of any given spectral 
character necessary to cause a sensation of brightness or 
brightness-difference. Under proper conditions the eye can 
function over a range of brightness of about io billion to 
one.30 Besides being a marvelous sensory organ the eye is 
also a motor organ capable of remarkably fine discrimina- 
tion. 
In considering the human eye as a sense-organ we cannot 
avoid giving weight to the conditions under which it evolved 
from a lowly sense-organ to the most highly developed 
one. It is not surprising to find this sense-organ, and the 
various processes which play parts in transferring its record 
to our consciousness, adapted best to daylighting conditions. 
It is not surprising to find the entire visual process better 
satisfied in general with the high intensities of illumination 
outdoors and with the quality of daylight. 
The remarkable range of sensibility of the eye is explained 
partially by the assumption of two different kinds of ele- 
ments in the retina and two corresponding visual regimes. FUNDAMENTALS OF VISION 
133 
For very low brightnesses retinal elements termed " rods " 
(owing to their form) are supposed to be in operation ac- 
cording to one theory which has much anatomical evidence 
to support it. They and their attendant processes enable 
us to see the brightnesses resulting from very low intensities 
of illumination such as those of moonlight, starlight, and 
feeble artificial light. The rods are relatively insensitive 
to hue or color as is evidenced by the colorless appearance 
of colored objects under low intensities of illumination. 
They do not exist in the central portion (fovea) of the 
retina; hence, we are unable to see a faint star by looking 
directly at it, although often we can see it by slightly averted 
vision. 
The other set of retinal elements are termed " cones " 
(owing to their form). These elements are supposed not to 
respond to feeble brightnesses. Their range of sensibility 
does not extend to as low brightnesses as the rods. They 
are responsible for at least most of our sensations of bright- 
ness at ordinary and higher levels of illumination and, if 
this " duplicity " theory is correct, they are primarily re- 
sponsible for practically all our color sensations. The cones 
predominate more toward the central portion of the retina 
while the converse is true of the distribution of so-called 
rods. 
Rods are absent from the fovea - a central region of the 
retina of approximately one degree of visual angle - and, 
in fact, are absent from the surrounding region having a 
total circular area of about 3 degrees. However, it should 
be pointed out that the cones in this central region are 
more rod-like structurally than the cones in the peripheral 
region. Cones and rods actually exist anatomically and 
there is evidence that they are the actual receptors or at 
least they play an important part in reception. Much of 
the data available point to them as performing the functions 
described above. Certainly the assumption of these two 134 
LIGHT AND WORK 
regimes explains much of the data pertaining to vision. The 
characteristics of vision over different parts of the retina 
seem to vary according to the distribution of rods and cones 
respectively. Incidentally, the smallest object that can be 
seen is that whose image falls upon only one retinal re- 
ceptor. This limits the resolving power of the eye if nothing 
else does. Normal eyes cannot readily resolve two objects 
nearer together than about 40 seconds of visual angle, which 
corresponds to a separation of about 0.0025 inch at a dis- 
tance of 14 inches from the eye. Data in Table XI and Fig. 
25 may be interesting at this point. 
It is thus seen that the eye or the visual sense-organ may 
operate in " low " or in " high." The change from one to 
the other is not as abrupt as that in the transmission of 
automobiles but the analogy is useful and it may be ex- 
tended even further. Just as the automobile is not designed 
to operate continually in " low " but only in emergencies 
requiring such a gear-ratio, so it is with the eye. For 
continuous operation the eye should be in " high." Many 
of the data presented in these chapters show that the eye 
does its best at high levels of illumination and their con- 
sequent brightnesses. At such brightnesses as those of the 
sun and of many artificial light-sources the eye or the visual 
process as a whole again becomes less sensitive. However, 
these brightness-levels need not bother us here because they 
are of much greater magnitude than even those of objects 
illuminated by direct noon sunlight. 
It will be noted that most of the curves in these chapters 
which show the increase in visual acuity, brightness sen- 
sibility, speed of vision, etc., are of the form of efficiency 
curves of electric motors, for example. It is not the general 
practice to operate motors on the lower parts of the curve 
which represent low efficiencies. It is good practice to 
operate them at the higher or highest efficiencies. So it is 
with the eye when economic considerations permit and, FUNDAMENTALS OF VISION 
135 
fortunately, the present efficiency of light-production and 
utilization make it economically possible to provide intensi- 
ties of illumination better suited to the operating charac- 
teristics of the eye than those prevailing indoors at the 
present time. In fact, it is shown in Chapter X that it is 
profitable to increase the generally existing intensities of 
illumination and to rehabilitate many lighting systems. 
Besides the high and low regimes in the visual process, 
we have the pupil of the eye as a governor to some degree. 
It performs the same function as the iris diaphragm in the 
camera for it regulates the average brightness of the retinal 
image. Its smallest aperture is about 2 millimeters (0.08 
inch) and its largest is about 8 millimeters (0.32 inch). 
This corresponds approximately to a range from F20 to 
about F2 on the usual aperture scale of a camera, the values 
being determined by dividing the focal length by the diam- 
eter of the aperture. It should be noted that the size 
of the aperture determines to some extent the definition of 
the image. The pupil has a similar influence on the retinal 
image. But when it is small, definition may be better, but 
the brightness of the retinal image of the same object under 
the same illumination is much less when the pupil is small 
than when the pupil is large. The resolving power of a 
lens increases as the aperture increases as is well evidenced 
by the desire for astronomical telescopes of large aperture, 
but apparently this influence is outweighed by other factors 
in the eye as shown later. 
We are all familiar with the results of adaptation of the 
visual sense to different levels of brightness or of illumina- 
tion intensity. Coming suddenly into an illuminated room 
from the dark we are " blinded " for some time until the 
eyes becomes adapted. This phenomenon is important in 
lighting, although lighting conditions to be best for work 
should be such that this phenomenon is not evident. 
Thus it is seen that there are many interesting and im- 136 
LIGHT AND WORK 
portant factors pertaining to the eye which have their in- 
fluence upon visual discrimination. Many others, particu- 
larly those involving color, could be discussed, but these 
would lead us far afield. Only the fundamental factors 
are considered here; the others may be found adequately 
treated elsewhere.1 
Illumination and Brightness of Test-Objects 
Many scientists have for years measured their visual 
test-objects in terms of brightness rather than illumination 
intensity. This is the correct and safe method. On the 
other hand the engineer and others are more familiar with 
foot-candles. Therefore, the diffuse reflection-factors of 
the different parts of the test-object should be given as well 
as the illumination in foot-candles. This makes it possible 
to compute the brightness. In these chapters the author 
has reduced many values to foot-candles on colorless surfaces 
diffusely reflecting, respectively, 80 per cent and 8 per 
cent of the incident light. This leaves the data in terms 
of foot-candles on so-called white and dark gray surfaces 
respectively in which form the data are most practicable. 
Test-objects are usually black on white. Of course, color 
plays an important role in vision and its influence should 
be studied in the laboratory; however, the black-white is 
the most natural combination. In general, this high con- 
trast will be a greater aid to vision than lower contrasts 
such as black on dark gray, and results with the black- 
white test-object will be representative more of how well 
the eye can do rather than of what it may be expected to do 
in every-day vision, excepting in work such as reading in 
which black-white contrasts are the rule rather than the 
exception. Therefore, it is important that low-contrast 
test-objects be also used in order to approach more nearly 
to average brightness-differences with which the worker is 
usually confronted. FUNDAMENTALS OF VISION 
137 
It should be noted that what we term black reflects some 
light and what we term white absorbs some light. There- 
fore, black is a very dark gray and white is a very light 
gray. Dull black paints and inks commonly reflect from 2 
to 5 per cent of the incident light; black felt about 1 per- 
cent; black velvet usually less than 1 per cent; and a hole 
in a box lined with black velvet emits practically no light. 
Black velvet surrounding the hole is bright compared with 
the hole. 
White papers and paints reflect from 70 per cent to 85 
per cent of the incident light. The purest magnesium car- 
bonate and similar white powders are sometimes found to 
reflect from 90 per cent to 98 per cent of the total light. 
However, a good white paint or paper usually has a diffuse 
reflection-factor of 80 per cent. 
Inasmuch as such a white is usually used, it is interesting 
to know its brightness under various intensities of illumina- 
tion. Brightness is measured in various units. Candles 
per square inch or the lambert are convenient units for the 
brightness of the sky and of artificial lighting units. The 
millilambert (one thousandth of a lambert) is a unit of 
convenient size for the brightnesses of objects under inten- 
sities of illumination up to 1000 foot-candles. Some of 
these relations are as follows: 
i lambert = 1000 millilamberts (ml) = 2.054 candles 
per square inch 
1 candle per sq. in. - 0.4868 lambert = 486.8 milli- 
lambert 
1 foot-candle on a perfectly white surface (100 per 
cent reflection-factor) produces a brightness of 1.076 
millilambert (ml) 
1 foot-candle on an ordinary " white " surface of 80 
per cent reflection factor produces a brightness of 
0.86 millilambert. LIGHT AND WORK 
138 
i millilambert = 0.929 lumen emitted per square 
foot - brightness that would be produced by 0.929 
foot-candles on a diffusing surface of 100 per cent 
reflection-factor. 
1 foot-candle = 1 lumen incident per square foot 
To find the brightness of a diffusely reflecting surface in 
millilamberts, the intensity of illumination in foot-candles 
is multiplied by the reflection-factor of the surface and this 
product is multiplied by 1.076. This relation is shown 
graphically in Fig. 54, Chapter XII. 
The original definition of the lambert (and the milli- 
lambert) involves perfect diffusion, so that the foregoing 
relations are best confined to surfaces that diffusely reflect 
or transmit light. In fact, throughout this book the term 
reflection-factor applies to diffusely reflected light unless 
otherwise stated. 
Brightness Sensation 
The most fundamental visual function is the distinction 
of brightness. It is the peculiar property of the visual sense- 
organ (and of its lines of communication to the brain and 
to our consciousness) that radiant energy between a certain 
range of wavelengths (about 390 mp. to 760 mp., Fig. 1 and 
Table I) impinging upon the retina causes the sensation of 
brightness. As is well known, radiation of various wave- 
lengths within this range causes sensations of certain hues. 
The combined sensation of radiations of various wavelengths 
is a result of a complex psycho-physiological process which 
need not concern us here. It has been seen in Chapter V 
that various spectral distributions of radiant energy give 
integral effects such as approximate white for noon sunlight 
and an unsaturated yellow for the light from tungsten fila- 
ment lamps. 
The minimum brightness of a given visual angle that we FUNDAMENTALS OF VISION 
139 
can see depends upon the state of adaptation of the eye 
and upon the wavelength of radiation. It is so very small 
for the dark adapted eye that it does not concern us in 
lighting although it is of scientific interest. In every-day 
vision the brightnesses which we encounter are much 
greater than this threshold value; therefore, in vision we 
are not so concerned with our ability to discern brightness 
as we are to distinguish brightness-difference. This visual 
function has been thoroughly studied by many investigators. 
Konig's data are presented in Fig. 22 for white light, and also 
MINIMUM PERCEPTIBLE BRIGHTNE55-DIFFERENCE 
100 
90 
60 
70 
60 
50 
40 
30 
20 
10 
0 
I 
(PERCENT) 
0.001 
0.01 
0.1 
1 
10 
100 
1000 
10000 
100000 
0.000001 
0.00001 
0.0001 
FOOT-CANDLES 
(QO°I° REFLECTION-factor) 
(logarithmic scale) 
Fig. 22. Showing the minimum brightness that is perceptible at 
various brightnesses corresponding to those produced by a wide 
range of intensities of illumination (foot-candles) of a so-called white 
surface of 80 per cent reflection-factor. The horizontal scale has 
been expressed in foot-candles on a colorless surface which diffusely 
reflected 80 per cent of the incident light as nearly as possible from 
the original data presented by Kbnig. The results for white light 
and also spectral blue and red lights are shown. 
for red and blue light, over a very great range of bright- 
nesses far exceeding the range ordinarily encountered in 
the objects that our work requires us to look at. The scale 
on the left represents the minimum brightness-difference 140 
LIGHT AND WORK 
(in per cent of the greater brightness) which is just percep- 
tible at the various brightnesses obtained by illuminating 
a so-called white surface (of 80 per cent reflection-factor) 
to intensities in foot-candles as shown on the horizontal 
scale. This scale had to be computed from data presented 
by Kbnig. The true sensitivity-curve of the eye to bright- 
ness-difference is obtained by plotting the reciprocals of 
the brightness-differences. The scale of illumination in 
Fig. 22 is also peculiar in that it is logarithmic. This en- 
ables the plotting of a great range of illumination intensi- 
ties (in reality brightnesses) because the scale is telescoped. 
By sensitivity is meant the ability of the eye to distinguish 
brightness-difference. 
In order to show the data in the usual manner the sen- 
sitivity of the visual sense to brightness-difference is plotted 
in Fig. 23 for a range of illumination from zero to 25 foot- 
candles on an ordinary white surface of 80 per cent reflec- 
tion-factor as shown in the lower scale of foot-candles. 
It shows, as does Fig. 22, that the eye can see much smaller 
differences in brightness if the brightnesses are those 
ordinarily encountered under high intensities of illumination 
than if they are of the lower values found under lower in- 
tensities of illumination. Here, as in many other cases, 
it is seen that the eye operates more effectively at high 
intensities of illumination than at low ones. It should be 
noted that brightness is the factor upon which vision largely 
depends. However, in practice we deal with foot-candles, 
so that, in these discussions and illustrations, foot-candle 
scales are provided, but the reflection-factor of the surface 
(or background in the case of visual acuity) is also given. 
This enables one to visualize and even to compute the 
brightness. 
It should be particularly noted that the brightness of a 
so-called white surface (80 per cent reflection-factor) is ten 
times greater than a so-called dark gray surface of 8 per FUNDAMENTALS OF VISION 
per cent reflection-factor under the same intensity of 
illumination. Much of our seeing is concerned with surfaces 
of low reflection-factor whose brightnesses are much lower 
than surfaces of high reflection-factor under the same in- 
141 
FOOT-CANDLES 
(8'1' REFLECTION-FACTOR) 
50 100 150 
60 - 
55 
50 
45 
40 
35 
30 
a 
20 
15 
10 
5 
ZOO 
Z50 
SENSITIVITY TO BRIGHTNESS-DIFFERENCE 
o 5 10 15 20 25 
FOOT-CANDLES 
(eo°lo reflection-factor) 
Fig. 23. Showing the sensitivity of the eye to brightness-difference throughout 
a range of brightnesses corresponding to illumination intensities up to 25 foot- 
candles on a so-called white surface of 80 per cent reflection-factor. The upper 
horizontal scale is the same as the lower one when both scales are reduced to 
brightness values because the upper scale is for foot-candles on a colorless surface 
that reflects only 8 per cent of the incident light. By sensitivity is meant the 
ability of the visual sense to see brightness-difference. 
tensity of illumination. The data in Fig. 23 are applicable 
to materials which reflect 8 per cent of the total light by 
using the upper scale of foot-candles. The curve now tells 
a different story for it is seen that the eye does not approach 
its maximum effectiveness within this range until the illumi- 
nation intensity has reached a value in the neighborhood of 142 
LIGHT AND WORK 
200 foot-candles. The two foot-candle values in Fig. 23 
represent at any vertical line, the respective intensities of 
illumination to produce equal brightnesses of two surfaces 
having 80 per cent and 8 per cent reflection-factors respec- 
tively. For a surface of any other reflection-factor the same 
curve applies if the foot-candle scale is changed accordingly. 
For example, if a surface of 40 per cent reflection-factor is 
considered, it is necessary merely to multiply the lower 
scale by 2 or to divide the upper scale by 5. This is a very 
practicable application of these data which has been gen- 
erally overlooked notwithstanding its importance in de- 
termining what effective lighting conditions really are. 
It should also be noted that the eye is called upon to 
distinguish brightness-differences not only of surfaces of 
low reflection-factors, but also when these are in shadows. 
In other words, all our vision is not confined to the desk, 
inspection table, planer-bed, etc. We must see under these, 
in the shadows on objects, and in the shadows cast by 
objects, etc. Here we may have low brightness-values due 
to the combination of reduced illumination and low reflec- 
tion-factors. The results in a shadow receiving only a 
certain per cent of the amount of light received by the 
work-plane can be determined for any surface of a given re- 
flection-factor by taking into account the fact that the 
illumination intensity is reduced to this certain percentage 
of the full illumination. These results are those obtained 
in the laboratory under ideal photometric conditions, where 
distraction is minimized and where the observer has nothing 
else to do at the time. The conditions are far different for 
the worker. Certainly he should have a large factor of 
safety; that is, the lighting should be such as to promote 
visibility, and the illumination intensity should be well up 
in the region of maximum visual efficiency not only for 
discrimination of brightness-differences of surfaces of high 
reflection-factor but also of low reflection-factor. Further- FUNDAMENTALS OF VISION 
143 
more, shadows should be taken into account. In other 
words, if the worker is to see quickly and with certainty he 
must have many times the amount of light necessary under 
FOOT - CANDLE5 
(©•|» REFLECTION-FACTOR) 
0 
240 ■ 
2Z0 ■ 
ZOO - 
180 
160 ■ 
140- 
IZO 
loo- 
se 1 
60 
40 
20 
50 100 150 ZOO 
Z50 
IJ Uli 
-hWh 
| _ 
' fe 
relative: sensitivity to brightness difference 
0 
5 
10 
15 
20 
25 
(BO'Io REFLECT!ON-FACTOR) 
FOOT- CANDLES 
Fig. 24. Showing the per cent increase in the sensitivity of the eye to bright- 
ness-difference for surfaces receiving full illumination (solid line) represented by 
the foot-candle scales and for a shadow (broken line) receiving only 10 per cent 
of this full illumination. The relative sensitivity is represented by 100 at the 
brightness of a colorless surface of 80 per cent reflection-factor under one foot- 
candle of illumination (or a surface of 8 per cent reflection-factor under 10 
foot-candles. 
the ideal conditions of laboratory researches. There should 
be a large factor of safety for the worker as discussed in 
Chapter XIII. 
It is interesting to ascertain the per cent gain in sensi- 
tivity of the eye to brightness-difference as the intensity 
of illumination increases. This is shown in Fig. 24 where 144 
LIGHT AND WORK 
ioo on the sensitivity scale represents the relative sensi- 
tivity of the eye to brightness-difference at the brightness 
of a colorless surface of 80 per cent reflection-factor under 
an illumination of one foot-candle (or one of 8 per cent 
reflection-factor under io foot-candles. The solid line 
represents the surfaces receiving the illumination represented 
in the foot-candle scales. It is seen that the sensitivity 
for brightness-differences on the so-called white surface, in- 
creases 75 per cent when the intensity of illumination is 
increased from one to io foot-candles (lower scale). This 
same increase in sensitivity at the same brightnesses, when 
surfaces of 8 per cent reflection-factor are involved, would 
be represented by an increase from io to ioo foot-candles 
(upper scale). 
It was pointed out that in every-day work it is often 
necessary to see details in shadows. These shadows are not 
only those cast by objects upon our work but are also those on 
the object itself. For example, we may find it desirable to 
see into complex machinery, under tables, etc., or we may 
find it necessary or desirable to inspect objects themselves 
without turning them in different directions so that all 
sides would eventually be seen under the full illumination. 
For this reason it is interesting to inquire into sensibility 
to brightness-difference in a shadow receiving only io per 
cent of the full illumination represented by the foot- 
candle scales in Fig. 24. This is shown by the broken line. 
It is seen that a 100 per cent increase in sensitivity (ability 
to distinguish brightness-difference) accompanies an increase 
in illumination from one to 10 foot-candles or from 10 to 
100 foot-candles respectively for the brightness represented 
by the so-called white surface or the so-called dark-gray 
surface respectively receiving the full illumination. This 
does not take into account the reduced sensitivity of the 
eye when looking into shadow by the influence of the higher 
brightness-level of surroundings receiving full illumination. FUNDAMENTALS OF VISION 
145 
All this argues for " luminous " shadows obtained by dif- 
fused and evenly distributed light and argues for higher 
intensities of illumination than are now generally used, if we 
are to operate the eye reasonably close to its highest 
efficiency. 
Many factors besides illumination intensity and bright- 
ness-level affect the sensitivity of the eye to brightness-dif- 
ference. The state of adaptation of the eye and the size 
of the stimulus are important31 in many cases of lighting 
where the intensity of illumination is far from uniform or 
of low values of the order of magnitude of moonlight or 
less. Reeves 32 has presented interesting data on the effect 
of visual angle and brightness of the stimulus (the object) 
on the ability of the eye to distinguish various brightness- 
differences after certain periods of time have elapsed since 
the eye was adapted to another brightness. In fact, a great 
amount of work has been done which involves the period 
and degree of adaptation. These data are exceedingly 
valuable in revealing the operation of the visual sense- 
organ. When we have adequate and proper lighting the 
eyes are not subjected to wide ranges of illumination or 
brightness and the time element of adaptation is not a 
prominent factor. Therefore, this phase will not be dis- 
cussed in this chapter. It is important to note that, other 
conditions being constant, the smaller the visual angle the 
greater is the contrast necessary to see an object. The 
data presented in Figs. 22, 23, and 24 are for large visual 
angles far above the limit of visibility. For small visual 
angles, the sensitivity of the eye to brightness-contrast 
would be less; hence, in those illustrations the foot-candle 
scales would have to be multiplied by factors greater than 
one to make the data applicable. In other words, for a 
small visual angle the intensity of illumination must be 
greater than for large ones if vision were to be equally 
effective in the two cases. 146 
LIGHT AND WORK 
Visual Angle 
An important factor in vision is the size of the object 
or details to be discriminated. Inasmuch as the size of the 
retinal image is more important than the size of the object, 
visual angle is the term in which acuteness of vision is de- 
termined. This takes into account both the size of object 
and its distance from the eye. Small visual angles are 
directly proportional to the size of the object divided by 
HEIGHT IN HUNDREDTHS DEAN INCH AT DISTANCE 0FI4 INCHES 
4000 
4 
RATIO OF DISTANCE TO SIZE 0E OBJECT 
% 
\ <p 
'x 
3000 
3 
ZOOO 
Z 
1000 
L-- 
I 
0 I 2345 6789 10 
VISUAL angle: 
(minutes or arc) 
o 
Fig. 25. Showing the relations of size of an object and its distance from 
the eye for small visual angles. The straight line represents the sizes of 
objects in hundredths of an inch (right-hand scale) at a distance of 14 inches 
from the eye for small visual angles. The curve shows the ratio of the dis- 
tance of an object from the eye to the size of the object in the same units 
of length (left-hand scale) for small visual angles. 
its distance from the eye. It is obvious that an object of 
a small size at one foot from the eye is of the same visual 
angle as an object of twice its size at a distance of two feet. 
Inasmuch as visual angles are so important, some data per- 
taining to small visual angles are presented in Table XI 
and in Fig. 25. It should be noted that in the second 
column of the table the size of the object is given in hun- FUNDAMENTALS OF VISION 
147 
dredths of an inch. For example, an object approximately 
2 hundredths of an inch in size and 14 inches from the eye 
subtends a visual angle of 5 minutes. The third column 
makes it possible to determine visual angle by size of object 
and distance from the eye in the same units of length. 
TABLE XI. 
The Relation of Size of Objects and Their Distance From 
the Eye for Small Visual Angles. 
Visual Angle 
(Minutes of arc) 
Size of Object 
r at distance of 14 inches 
(in hundredths of an inch) 
Distance of 
object divided 
by size of 
object in same 
unit in length 
i 
0.406 
3448 
2 
0.812 
1724 
3 
1.218 
ii49 
4 
1.624 
862 
5 
2.030 
689 
6 
2.450 
57i 
7 
2.856 
490 
8 
3.262 
429 
9 
3.668 
382 
10 
4-074 
344 
Visual Acuity 
After the ability of the retina to distinguish brightness 
and brightness-difference, the next important factor is 
acuteness of vision. We distinguish a certain minimum 
brightness-difference between an object and its background 
and naturally we are curious to know how small the object 
would have to be to be undistinguishable, other factors re- 
maining constant. The term visual acuity is applied to the 
discrimination of fine detail. It is usually determined as 
the smallest visual angle between two objects which the 
eye can resolve under the particular conditions. Some idea 
of the relation of size and distance for various visual angles 
may be gained from Table XI and Fig. 25. 
A test-object for the study of visual acuity may be any 
fine detail such as letters, fine lines, broken circles, dots, 
etc. The visual angle is determined by measuring the di- 
mension of the part to be discriminated and the distance 148 
LIGHT AND WORK 
to the eye. The visual angle may be varied by moving the 
object toward and from the eye, but this is not a satisfac- 
tory method for analytical work. It introduces factors such 
as change of focus. In such tests it is well to have such an 
object as parallel bars or a split ring which can be rotated 
about its center so that the observer may be required to 
state the position of the bars or missing part of the ring, 
thereby showing whether or not he actually discriminated 
the bars or two parts of the ring. Some features of test- 
objects have been discussed in a previous section. 
, FOOT - CANDLES 
(6*/« reflection-factor) 
25 
AO 
60 
IZO 
160 
ZOO 
240 
280 
320 
c-Mormol eqs . 
1 Jr - -L - 
'tAstigmatic eqe 
2.0 
VISUAL ACUITY 
1.5 
1.0 
0.5 
0 
4 
8 
12 
16 
20 
24 
28 
32 
FOOT-CAN DLLS 
(60°|o REFLECTION-FACTOR) 
Fig. 26. The solid line represents the general relation between visual 
acuity and illumination intensity for a normal eye, the test-object con- 
sisting of black details on a so-called white background of 80 per cent 
reflection-factor (lower scale). The upper scale applies to a dark gray 
background of 8 per cent reflection-factor. The dotted line shows the 
general relation of the results obtained with a slightly astigmatic eye. 
In Fig. 26 the solid line represents for the normal eye 
the typical relation between visual acuity and the illumi- 
nation (lower scale) of a " black " test-object on a " white " 
background of 80 per cent reflection-factor. It is seen that 
the form of this curve is quite similar to that of many 
pertaining to vision which are represented in this and other FUNDAMENTALS OF VISION 
149 
chapters. It is the familiar " efficiency curve." Its actual 
form depends somewhat upon the character of the test- 
object but that is a matter more of scientific than of prac- 
tical interest. The solid line in Fig. 26 is representative 
of the general form of the visual-acuity curve or general 
relation of visual acuity and brightness of background of 
black details. 
At low levels of illumination or brightness of background, 
visual acuity is much less than at higher levels. It should 
be noted that these results are for the ideal laboratory 
conditions and for a test-object of maximum contrast and 
white background. Assuming that the black of the test- 
object is non-reflecting and that the background is a dark 
gray of 8 per cent reflection-factor, we now must apply the 
upper scale of foot-candles. For a surface of 40 per cent 
reflection-factor the lower scale would be multiplied by 2 
or the upper one would be divided by 5. A foot-candle 
scale can be thus constructed for any other reflection-fac- 
tor. We see now that the eye even under the ideal con- 
ditions of the laboratory does not begin to operate any 
where near its maximum effectiveness in this respect until 
the illumination of the test-object is above 10 foot-candles. 
If we consider the matter of a factor of safety to allow 
for the less ideal conditions in every-day work and perhaps 
for discriminating details in shadows, it is seen that illumi- 
nation intensity from the viewpoint of visual acuity should 
be a rather high value compared with one foot-candle for 
example. This is discussed further in Chapter XI. 
It is also of interest to ascertain the per cent increase in 
visual acuity with increasing brightness or intensity of 
illumination on the two colorless surfaces reflecting respec- 
tively 80 per cent and 8 per cent of the total incident light. 
In Fig. 27 for the sake of comparison the relative visual 
acuity is represented by 100 at a brightness corresponding 
to one foot-candle on an 80 per cent reflecting (colorless) 150 
LIGHT AND WORK 
surface, or io foot-candles on the 8 per cent reflecting 
(colorless) surface. As already stated, all our seeing is 
not done on objects and surfaces receiving full illumina- 
tion; that is, often we must discriminate details in shadow. 
FOOT- CANDLE5 
(8°l» REFLECTION - FACTOR) 
200 
40 60 IZO 160 ZOO 240 280 5Z0 
UK 
-A \0>" 
175 
150 
PERCENT INCREASE IN VISUAL ACUITY 
IZ5 
100 
75 
50 
25 
0 
4 
© 
IZ 
16 
20 
24 
29 
3Z 
FOOT-CANDLE J 
(80 /•REFLECTION--FACTOR) 
Fig. 27. The solid line shows the per cent increase in visual acuity 
with brightness of background against which the black details are viewed. 
The foot-candle scales apply to surfaces of 80 per cent and 8 per cent 
reflection-factors respectively. The relative visual acuity is represented 
by 100 for a brightness of background corresponding to one foot-candle 
on a colorless surface of 80 per cent reflection-factor (or io foot-candles 
on a colorless surface of 8 per cent reflection-factor). The broken line 
shows the per cent increase of visual acuity in a shadow receiving only 
io per cent of the full illumination represented by the foot-candle scales. 
Therefore, the per cent increase in visual acuity is also 
shown for brightnesses of shadows receiving only io per cent 
of the full illumination represented by the foot-candle 
scales. For surfaces having reflection-factors other than 
80 per cent or 8 per cent, the foot-candle scale can be altered FUNDAMENTALS OF VISION 
151 
to fit. For example, for a surface of 40 per cent reflection- 
factor the lower foot-candle scale would be multiplied by 
2 or the upper one would be divided by 5. 
The foregoing discussion applies solely to normal eyes. 
Of course all eyes that have defects which can be counter- 
acted by proper glasses should be equipped with such glasses. 
Nevertheless many persons do not wear glasses, notwith- 
standing the fact that their eyes need slightly corrective 
glasses at least. Progressive industries have found that 
examination of the eyes of workers pays them as well as 
the workers; nevertheless, in lighting it is well to consider 
not only the requirements of normal eyes but of some- 
what defective ones. If a reasonable factor of safety is 
allowed for these, subnormal eyes have less of a handicap 
than if there is no allowance for them. The reduced visual 
acuity of eyes slightly astigmatic is shown in general by the 
broken line in Fig. 26. In some industries as many as 50 
per cent of the workers have been found to have defective 
vision. 
It is a fact of common observation that we can see fine 
details of ordinary brightnesses or under ordinary intensi- 
ties of illumination can see exceedingly more clearly if we 
look directly at the details than if we look a little to one 
side of them. The converse is true for abnormally low 
brightnesses or illuminations. This is well shown in Fig. 
28. The broken line shows relative visual acuity for various 
parts of the retina at levels of illumination under which 
the eyes are generally used for fine work. The solid line 
represents the visual acuity under very low intensities of 
illumination. Zero degrees represents the center of the ret- 
ina, that is, the center of the fovea. It is seen that for 
the high-level brightnesses or illuminations visual acuity is 
very high at the center of the retina but it drops very 
rapidly in value for the first few degrees. For the abnor- 
mally low-level brightnesses or illuminations, such as moon- 152 
LIGHT AND WORK 
light and starlight, the converse is true. The central part 
of the retina is not as effective as an area between io and 
20 degrees from the axis or center. This shows why we 
see a faint star better by slightly averted vision. 
160 - 
160' 
140 ■ 
JZO 
100 
60 ■ 
60- 
40' 
2.0- 
RELATIVE VISUAL ACUITY 
Blind spot 
60 40 Z0 0 to 40 60 
DEGREES FROM OPTICAL AXIS 
Fig. 28. Showing relative visual acuity for various 
portions of the retina. Zero degrees is at the center of 
the fovea of the retina. The dotted line represents the 
relative visual acuity at ordinary brightnesses of high- 
intensity illumination. The solid line represents visual 
acuity for very low brightnesses (low intensities). 
The blind-spot in the eye is shown by the gap in the two 
lines to the right of the axis. This is of no importance 
excepting in monocular vision because the blind-spots of 
the two eyes do not overlap in the visual field. The blind- 
spot can be detected by fixing one eye (the other being 
closed) on a black dot on a white surface and fixing the 
attention on another dot on the same surface about two 
inches from the first. If the white surface be held about 
a foot from the eye and moved back and forth, a point will 
be found where the second dot will disappear. 
Color Sensation 
At ordinary levels of illumination intensity (or of bright- 
ness), we can see about the same brightness-difference of FUNDAMENTALS OF VISION 
colored surfaces as of colorless surfaces. Under the best 
laboratory conditions this brightness-difference is about 1.6 
per cent as seen in Fig. 22. Thus between a perfect white 
and a perfect black we have about 60 perceptible shades 
for any given high intensity of illumination. This range 
is not as great with colored surfaces because, for example, 
a fairly pure red paint would reflect only about 20 per cent 
of the incident light. Thus between it and a perfect black 
we would have only about 12 perceptible shades. 
A tint is a color diluted with white light (or pigment). 
The reflection-factors of tints of a given color vary from 
that of the original color to the limiting tint which is white 
itself. Therefore, the number of distinguishable shades 
of a tint depends upon the brightness of the tint itself. 
An approximate estimate of the average number of dis- 
tinguishable shades of white, of tints, and of colors might 
be taken as 30 or half the number of neutral grays dis- 
tinguishable between black and white. This is for a given 
illumination intensity of reasonably high value. A shade 
of any color (including tints and even white) is merely a 
lesser brightness. Therefore, we can produce shades in the 
broad sense by increasing the intensity of illumination. 
If we have a so-called white under a given intensity of 
illumination, by increasing the intensity of illumination we 
produce a lighter shade (in an absolute sense) than the 
original white. Therefore, the number of absolute shades 
which we can see is very much greater for any color, in 
general, than the series of distinguishable shades under any 
given intensity of illumination. 
With sensitive apparatus we can distinguish about 125 
hues in the spectrum of sunlight or light from a tungsten 
filament. (In the light from the mercury-arc there are 
only four principle hues and a few others represented by 
very weak spectral lines.) Considering that the purples 
are not represented in the spectrum, we may estimate that 
under the best laboratory conditions we can see 150 dis- 
153 154 
LIGHT AND WORK 
tinctly different hues. Our ability to distinguish a dif- 
ference in tint (technically termed difference in saturation) 
is not as well developed as our ability to distinguish dif- 
ferences in hue or in brightness. Data on large groups of 
untrained eyes are not available but we will estimate that 
20 different tints (or degrees of saturation) represent the 
average number of tints of a color actually distinguishable. 
Therefore, by multiplying the number of distinguishable hues 
(150) by the estimated average of distinguishable shades 
(30) under a given high intensity of illumination by 
the estimated average of distinguishable tints (20) we 
obtain 90,000 as the approximate number of different color 
sensations that we can distinguish for a given high intensity 
of illumination. When we introduce absolute brightness, 
that is, increase the intensity of illumination and thereby 
introduce an extensive range of brightness, it is seen that 
the number of distinctly different visual sensations which 
we can experience (apart from those of form) runs into 
millions. 
P. G. Nutting and L. A. Jones found that the sensibility 
of the eye to change in tint varied with the depth of tint. 
That is, to a pure color as much as 4.7 per cent of white 
light had to be added before the change in tint was noticed. 
After the color had been diluted so that it was a medium 
tint (50 per cent pure hue and 50 per cent white) the addi- 
tion of 4 per cent of white light was just noticeable as a 
change in tint. As the color became less and less saturated 
by the addition of more and more white, this percentage, 
required to produce a just noticeable tint, decreased until 
it reached a value of 2.2 for the very lightest tints (those 
very nearly colorless),. A pure color is said to be of 100 
per cent saturation. Tints are unsaturated colors of various 
percentages of saturation. 
In color-perception at very low levels of illumination the 
Purkinje effect must be taken into account and also the FUNDAMENTALS OF VISION 
155 
difference in our sensibility to brightness-difference for 
various colors. 
In very fine discrimination of color, after-images or suc- 
cessive contrast must be taken into account. The portion 
of the retina fatigued to a certain color will have an after- 
image of the complementary color. 
A central region of the retina is called the " yellow spot " 
because it absorbs blue rays somewhat and therefore in this 
small portion of the visual field colors in general are not 
seen as they are. This spot can be noted sometimes if 
the eyes are fatigued by bright colors. It is also noticed 
on viewing shades of colors of peculiar spectral charac- 
teristics. 
For accurate color-discrimination an intensity of at least 
50 foot-candles is more generally desirable than lower in- 
tensities often encountered. 
Simultaneous Contrast 
If we juxtapose two quite different colors, in general, 
they will both appear more vivid and different than they 
will under other conditions. If we place a gray paper on a 
green background of the same brightness, the gray paper 
will assume a purplish tinge complementary to the sur- 
roundings. This effect is the life of color and is in opera- 
tion everywhere we have color-difference. It can be demon- 
strated on any colored painting or print. Suppose we have 
an orange flower-pot amid green foliage. The orange may 
appear quite vivid under these conditions. Punch a hole 
in a white or gray paper and place the hole over the orange 
pot thereby excluding the other colors. The orange now 
has lost much of its colorfulness. In the photographic dark- 
room we gradually become accustomed to the red light and 
it loses much of its redness. If we discover a crack letting 
in daylight, the daylight appears a blue-green color. If we 156 
LIGHT AND WORK 
light a lamp of any other color than red, the red light im- 
mediately assumes its redness again. 
The effect of simultaneous contrast is also very notice- 
able with colorless surfaces. A white area on a dark gray 
appears much brighter than it is. Conversely the dark 
gray appears darker than it is. This is very prominent in 
a series of grays. Where a brighter edge meets a darker 
one the former will appear brighter along the edge than 
further away from the edge. The converse is true of the 
dark edge. 
We can never see a color as it actually is because we can 
not escape the influence of its environment. 
Irradiation 
If we cut a hole out of a white card and place it over 
a black card the black hole appears smaller than it really 
is. Conversely a white area of the same size but placed on 
a black surface will appear larger than it is. When we 
turn on a light-source, particularly in a diffusing globe, it 
appears to move slightly toward the observer. This is an 
interpretation of apparent increase in size (due to irradia- 
tion) into an apparent approach. The most striking illus- 
tration of irradiation is a glowing fine filament of an electric 
lamp. It appears many times larger than the actual fila- 
ment. Even at a low voltage when the fine filament is 
barely glowing it appears considerably larger than it actually 
is. It is interesting to note how the size appears to increase 
from the cold junction to the fully glowing portion. Irra- 
diation is a very important phenomenon particularly in 
connection with backgrounds as discussed in a later para- 
graph. 
The Pupil of the Eye 
A brief description of the eye has already been given 
and it is illustrated in Fig. 21. It is seen that the pupil FUNDAMENTALS OF VISION 
157 
offers some protection to the eye against sudden changes 
in brightness or in illumination levels. If its minimum diam- 
eter approaches one millimeter and its maximum about 8 
mm., its area at the largest diameter is more than 50 times 
that at its smallest diameter. Inasmuch as the brightness 
of the image focused upon the retina is proportional to the 
area of the pupillary aperture, this brightness is under 
control of the pupil over a range of about 50 times. Com- 
bining with this control the peculiar property of " adapta- 
tion of the eye," we have a sense-organ that is fairly well 
fitted for extensive ranges in brightness. 
As already shown, the pupil plays important optical roles 
but some of these are at cross purposes. For example, the 
definition of the image is best for low apertures (small 
pupils) but the resolving power of any lens increases with 
the diameter of the aperture. However, definition is more 
important than resolving power in the optical parts of the 
eye because the retinal structure limits resolving power. 
The pupil of the eye under ordinary conditions of lighting 
is contracted considerably. Apparently its function is 
chiefly to enlarge at low intensities of illumination to admit 
more light even at a sacrifice of definition. The area of the 
pupil having a great deal to do with the brightness of the 
retinal image, it is important in vision. Inasmuch as 
glaring light-sources tend to make the pupil contract we not 
only have a reduced visibility due to glare but also due to a 
decreased brightness of the retinal image. 
It is easy to study the effect of pupillary aperture by 
using artificial pupils close to the eye. A series of circular 
openings of different sizes can be made in opaque material 
like sheets of thin metal. Dr. P. W. Cobb 33 studied the 
effect of various pupillary apertures on visual acuity. The 
test-object consisted of fine black lines on a white back- 
ground. The width of the lines and the intervals between 
them were equal and could be varied in number per inch with- 158 
LIGHT AND WORK 
out disturbing the equality of width of lines and of the 
intervals between them. In Fig. 29 Dr. Cobb's data on the 
relative visual acuity of the eye for various pupillary aper- 
tures are shown in the solid line A. The actual visual 
angles between the center of a black line and the center 
of the bright interval are given on the right-hand scale. 
6 
0.54 
7 
A 
0.6Z 
6 
.5. 
O.7Z- 
5 
VISUAL ANGLE IN MINUTES 0E ARC 
0.86 
VISUAL ACUITY 
(RELATIVE) 
4 
1.05 
3 
1.44 
2 
Z.I6 
I 
4.32 
0 
I 
DIAMETER OF ARTIFICIAL PUPIL 
(MILLIMETERS) 
Z 
3 
4 
5 
6 
Fig 29. The effect of pupillary aperture on visual acuity. The 
full line A represents results obtained with various apertures and a 
test-object of constant brightness. The broken line B represents 
results obtained with various pupils but with the brightness of the 
test-object altered in each case to keep the retinal image of constant 
brightness. The brightness of the test-object was the same for the 
1 mm. pupil in the two cases. 
It will be noted that this is not a uniform scale. Inciden- 
tally the product of the relative visual acuity and the 
corresponding visual angle is equal to a constant through- 
out the two scales. It is seen that visual acuity reaches a 
maximum at a pupillary aperture of about 4 mm. The FUNDAMENTALS OF VISION 
159 
brightness of the test-object was 189 candles per square 
meter. However, it is interesting to note that here we have 
several variables due to the variation in pupil size. As the 
pupil increases in size, definition decreases but resolving 
power and brightness of retinal image increase. It has been 
seen that visual acuity increases as the illumination inten- 
sity of the test-object increases. According to Dr. Cobb's 
data the best size of the pupil from the viewpoint of visual 
acuity, external conditions being constant, is from 2.5 mm. 
to 4.5 mm. The diameter of the pupil under satisfactory 
conditions of lighting is usually in the neighborhood of 3 
mm. to 4 mm. 
It is interesting to note that over the range of artificial 
pupillary apertures from 1 mm. to 5.6 mm., the brightness 
of the retinal image increased over 31 times for the data 
represented by the solid line A in Fig. 29, although the 
brightness of the test-object was constant at 189 candles 
per square meter. Dr. Cobb then conducted the interesting 
experiment of keeping the brightness of the retinal image 
constant by decreasing the illumination intensity of the 
test-object in proper proportion to the increase in pupillary 
aperture. In other words, when the largest artificial pupil, 
5.6 mm., was used the intensity of illumination at the test- 
object was less than one thirty-first of its value when the 
1 mm. pupil was used. The data obtained in this experi- 
ment are shown by the broken line B. When the pupil was 
1 mm. in diameter, the brightness of the retinal image was 
the same for each case but the lower visual acuity shows 
the result of the low constant brightness of the retinal image 
compared with the continually increasing brightness of the 
retinal image in the case of the data represented by the 
solid line. This shows that for a constant brightness of the 
image on the retina the size of the pupil has little effect 
on our ability to see fine detail within the range from 2 mm. 
to 5.6 mm. 160 
LIGHT AND WORK 
This is exceedingly interesting inasmuch as we need not 
consider the effect of the size of the pupil on visual acuity 
excepting as it influences the brightness of the retinal image. 
The latter is a very important factor in lighting because an 
increase in brightness of the retinal image by an increase in 
pupillary aperture is just as beneficial to vision as an 
equivalent increase in illumination intensity when the pupil 
remains constant in size. The data show that the pupil 
size does not appreciably alter the standing of the eye as 
an optical device for seeing fine detail within the range 
from 2 mm. to about 5 mm. in diameter. This represents 
a range of brightness of the retinal image of about 6 times. 
If the pupil could be kept open to a diameter of 5 mm. 
instead of 2.5 mm. under a certain condition, we would 
have visual results equivalent to a 300 per cent increase 
in intensity of illumination. No better argument can be 
devised against unshaded light-sources, excessive brightness- 
contrasts and specular reflection of bright images, all of 
which are glaring and cause the pupil to contract. Well 
diffused light, properly controlled and distributed, tends to 
open the pupil and thereby increase the effectiveness of 
the illumination. It should be noted that even though the 
pupil may contract upon being exposed to higher bright- 
nesses it also tends to become larger as adaptation to the 
new condition progresses. Reeves 34 has shown that after 
a period of 15 minutes' adaptation to darkness, the average 
size of the pupil for 6 different persons was 8 mm. He 
repeated the experiment for a number of brightnesses up to 
2000 millilamberts (about 4 candles per sq. in.). The 
average size of the pupil after 15 minutes' exposure to this 
highest brightness was 2 mm. His subjects were less than 
middle age. Measurements of the pupil externally yield 
results slightly greater than the true size of the pupil and, 
as Reeves pointed out, his values should be decreased 
by about 7 per cent. In Fig. 30 his data, averaged FUNDAMENTALS OF VISION 
161 
for six subjects, are plotted to a logarithmic scale of 
brightness (foot-candles on a surface of 80 per cent reflec- 
tion-factor) in order to cover an extensive range of bright- 
ness. It should be pointed out that the brightnesses of light- 
ing units are ofttimes more than ten times the maximum 
brightness used by Reeves, and exposed light-sources and 
0 
7 
6 
5 
4 
3 
2 
I 
0 
DIAMETER OF PUPIL IN MILLIMETERS 
100000' 
10000' 
1000' 
IOO' 
.01 
I 
I 
10 
100 
1000 
10000 
FOOT-CANDLES 
(60*[» REFLECTION FACTOR) 
(LOGARITHMIC SCALE) 
Fig. 30. Showing the size of the pupil after 15 minutes' exposure to 
various brightnesses from zero to 2000 millilamberts (about 4 candles per 
sq. in.). 
their specularly reflected images are thousands of times 
greater than his maximum brightness. However, these 
results are extremely interesting in connection with bright- 
nesses encountered in connection with everyday work under 
proper conditions of lighting. They show what we may 
expect of the pupil of the eye. The maximum size of the 
pupil varied for the different subjects from 7 mm. to 8.7 mm. 
At 100 ml. the variation was from 2.7 mm. to 2.9 mm. 
In Fig. 31 the full line represents Reeves' data on the 
closing of the pupil after the stimulus was suddenly changed 162 
LIGHT AND WORK 
from darkness to 100 millilamberts. These data are the 
average for six observers. The broken line represents his 
data on the opening of the pupil after the stimulus, which 
was a brightness of 100 millilamberts, was suddenly re- 
placed by total darkness. It is interesting to note that the 
RATE Or OPENING OF THE PUPIL 
(SECONDS) 
8 
40 
80 
IZO 
160 
ZOO 
240 
280 
320 
Opemngjoupil 
7 
6 
DIAMETER OF PUPIL IN MILLIMETERS 
5 
4 
3 
Closing pupil 
2 
I 
0 
4 
6 
12 
16 
20 
24 
2d 
32 
RATE OF CLO5IN6 OF THE PUPIL 
(JECOND5) 
Fig. 31. Showing the rate of opening the pupil after a brightness of 
100 millilamberts has been replaced by darkness; and the rate of closing 
after darkness has been replaced by the stimulus of 100 millilamberts. 
pupil requires about as many minutes to open to its maxi- 
mum diameter as it does seconds to close to its minimum 
for the conditions of the experiment. Doubtless, the rate 
of opening would be increased with very powerful stimuli 
and possibly the rate of closing would be slower; neverthe- 
less, the great difference between the rate of opening and of 
closing would not be wiped out. It is such data as these FUNDAMENTALS OF VISION 
163 
that help us to understand visual results and to predict 
what we may expect under given lighting conditions. 
Adaptation of the Eye 
This phenomenon is well known to everyone by experience 
and, doubtless, everyone knows that it takes time for the 
eye to adapt itself to changes in brightness. The pupil tries 
to do its part by changing in size, but there is still a more 
important adaptation which is retinal or perhaps psycho- 
physiological. Sometimes in lighting, the question of quanti- 
tative data arises; therefore, a brief discussion appears 
worth while. On coming into a dark room from outdoors 
soooo 
70000 
60000 
50000 
40000 
30000 
ZOOOO 
10000 
SENSITIVITY 
0 10 20 30 40 50 60 
TIME IN DARK 
(minutes) 
Fig. 32. The increase in sensitivity of the eye with time of 
exposure to darkness as determined by the minimum brightness 
perceptible as the time of adaptation to darkness increased. 
in the daytime it takes a long time for the eye to adjust 
itself. The room may not be perfectly dark, but it may 
seem so for several minutes. The lowest brightness that 
can be seen at first is many times brighter than the lowest 
perceptible brightness after adaptation to darkness for ten 
minutes. The sensitivity of the eye does not reach its 164 
LIGHT AND WORK 
highest point until it has been adapted to complete darkness 
for about an hour. Many have studied the increase in 
sensitivity of the eye as the period of adaptation decreases. 
In Fig. 32 the data obtained by Hecht35 have been plotted. 
It is seen that the sensitivity of the eye increases very slowly 
for 5 or 10 minutes and then increases rapidly for the 
period of 10 to 30 minutes after entrance into a perfectly 
dark room. It continues to increase in sensitivity up to 
about 50 minutes. During this entire period the sensitivity 
has increased over 70,000 times. It may appear from Fig. 
32 that practically no change occurs in the sensitivity of 
the eye to brightness in the early stage of adaptation to 
darkness, but it should be noted that the large values of the 
left-hand scale obscure the small values of the first few 
minutes. In fact, during the first five minutes of dark 
adaptation, the least brightness that the eye can see drops 
to about 50 per cent of its initial value. 
The slow increase for the first few minutes shows the un- 
desirability of subjecting workers to various levels of 
illumination intensity for short periods of time. The eye 
cannot adjust itself rapidly enough to radically different 
brightnesses to justify tolerance of such conditions even 
in the same room, if workers are to go to and fro from one 
level of illumination to another. Furthermore, great bright- 
ness-differences should not be tolerated in the visual field. 
Glare from unshaded light-sources, automobile headlamps, 
and street-lamps is very much a matter of the state of adapta- 
tion of the eye. A lighted match may be glaring to a dark- 
adapted eye but an electric filament lamp having a bright- 
ness and a luminous intensity much greater can be viewed 
without as much discomfort against the bright sky by eyes 
adapted to outdoor daylight. An eye adapted to outdoor 
daylight will tolerate brightnesses thousands of times 
greater than a dark-adapted eye. The subject is too com- 
plex to discuss in detail here but it may be stated as a FUNDAMENTALS OF VISION 
165 
fundamental of good lighting conditions, that the eyes should 
not be subjected to sudden changes in brightness or to sud- 
den changes in general levels of illumination of great 
magnitudes. 
The Purkinje Effect 
Visible radiations of the shorter wavelengths (violet, blue, 
green) are relatively more effective in causing brightness- 
sensation at low intensities than at high ones. The converse 
is true of visible radiations of longer wavelengths (yellow, 
orange, red). This phenomenon, discovered by Purkinje 
nearly a century ago, is of great scientific interest and has 
some importance in lighting at low intensities of illumina- 
tion. It is seen in Fig. 22 that the threshold value of blue 
light is much less than that for red light. This suggests the 
Purkinje effect. Suppose we equally illuminate a white 
surface with blue light and one with red light to a bright- 
ness equivalent to that of a white surface of 80 per cent 
reflection-factor under an illumination of one foot-candle of 
white light. Now if we have some means of reducing the 
intensity of illumination at exactly the same rate for the 
blue and red lights, when the brightness becomes equivalent 
to the white surface under 0.2 foot-candles of white light, 
the blue will begin to appear brighter than the red. If 
we continue the reduction, the blue becomes relatively 
brighter and brighter. Finally, the point will be reached 
(equivalent to the white surface under 0.001 foot-candle 
of white light) when the red will be practically invisible 
but the blue will still be perceptible. If we had increased the 
intensities instead of decreasing them no such effect would 
have been found. Perhaps the red would have become 
slightly brighter than the blue but not very much brighter. 
Certainly Konig's data or any data that the author is 
familiar with do not indicate that this would be true. 
It is thus seen that an illuminant that is bluish in com- 166 
LIGHT AND WORK 
parison with another would be somewhat more effective at 
very low intensities than the yellowish one. This advan- 
tage may be appreciable at intensities of illumination of the 
order of magnitude of 0.05 foot-candles or less but it is 
negligible for any commercial illuminants at intensities 
above 0.1 foot-candle. In signalling and wherever colors 
of very low brightness are seen by eyes adapted to low 
brightnesses, this ' phenomenon may be worth taking into 
account, but, excepting for very low intensity lighting, it is 
of little practical interest in general lighting practice. 
Often claims are made in factory lighting that a yellowish 
illuminant is not as satisfactory as a white or a bluish one 
for visibility in the shadows. This is based on the assump- 
tion that the Purkinje effect is operative in the shadows 
while the eye is adapted to high intensities at the work and 
elsewhere. This assumption is dangerous when it is not 
supported by reason or by data. The Purkinje phenomenon 
belongs to the " low " regime of the visual process which can 
not be appreciably effective when the eye is adapted to 
high levels of illumination. Furthermore, it has been shown 
the the Purkinje effect is not perceptible on a portion of 
the retina when other parts of the retina are subjected to 
ordinary brightnesses. 
It is also of interest to note that there is no evidence of 
the Purkinje effect in the central portion of the retina 
over an area at least 2 degrees of visual angle. 
Growth and Decay of Visual Sensations 
A sensation of brightness or of color does not grow im- 
mediately to its full or final value the instant the image is 
focused on the retina, and it does not immediately decrease 
to zero when the stimulus disappears. In fact, the bright- 
ness-sensation due to a certain stimulus " overshoots " at 
first if the stimulus is bright enough. Doubtless many per- 
sons have noticed this in turning on the electric filament 
lamps in a room. Those who have not may be interested FUNDAMENTALS OF VISION 
167 
to note the slight " overshooting " of general brightness of 
the ceiling, for example, an instant after the switch is 
pushed. It is more readily noticed if the lighting is of un- 
usually high intensity. This point is not of general interest 
but it arises in some of the more scientific phases of lighting. 
7.5 
7.0 
6.5 
6.0 
5.5 
5.0 
4.5 
4.0 
as 
ao 
2.5 
2.0 
1.5 
1.0 
0.5 
STEADY ILLUMINATION OFWHITE SCREEN IN (APPROX.) FOOT-CANDLES 
ZZ Foot-candle<, 
22 
3 
6 
1.6 
0 0.05 0.1 0.15 O.Z 0.Z5 0.3 0.35 
TIME IN SECONDS 
Fig. 33. Illustrating the overshooting of brightness sensation. The 
figures are foot-candles and the brightnesses are those produced by these 
illumination intensities on a so-called white surface of 80 per cent reflection- 
factor. 
At any rate, it is of some interest in connection with the 
operation of the visual sense-organ. It can be demonstrated 
by means of specially constructed rotating disks and in 
other ways. 
The phenomenon is shown in Fig. 33 by means of data 
obtained by Broca and Sulzer.37 They compared the bright- 168 
LIGHT AND WORK 
ness of a so-called white screen under a certain intensity 
of illumination with a steady brightness at various intervals 
after the illumination was applied to the white screen. They 
used five different brightnesses obtained by illuminating 
the white surface to five different intensities of illumination. 
The brightnesses were those produced by illuminating a 
so-called white surface of 80 per cent reflection-factor to 
the intensities in foot-candles indicated by the figures. It 
is seen that the sensations due to the higher brightnesses 
" overshot " about 0.05 second after the stimulus was 
applied. When the stimulus was of lower brightnesses the 
overshooting was not marked. 
The author 36 and others have studied the overshooting 
of color-sensations. In the case of signals of short dura- 
tion this overshooting is of real value. In general, the 
hues in the mid-spectral region overshoot less than the others. 
Blue overshoots more than red and red more than green. 
After-images are evidence that sensations do not decay 
to zero immediately upon the cessation of the stimulus. 
Usually the color of after-images of a colored stimulus are 
complementary to that of the original stimulus, excepting 
for the early period of decay of extremely powerful sensa- 
tions. Such after-images are to be reckoned with in many 
visual problems. Incidentally, the persistence of vision is 
the basis of motion-pictures. 
After-images of bright objects, and particularly of light- 
sources, are not only annoying but often are temporarily 
blinding. Many accidents are caused by glaring light- 
sources in factories and on the streets. When not actually 
a source of danger they reduce the ability to see. This is 
one of the powerful arguments against glaring or unshaded 
light-sources in lighting systems. The author attempted 
to measure the duration of after-images by measuring the 
time it took them practically to disappear. The data are 
presented in Fig. 34. The highest brightness employed was FUNDAMENTALS OF VISION 
169 
that of the brightest electric filament lamp at the time. 
This was the vacuum tungsten lamp and the brightness of 
its filament was 1080 candles per sq. in. The duration 
of the after-image increased with the time of exposure to 
the bright filament. In the case of the highest brightness 
of filament, looking at the filament for one second caused 
an after-image that endured for 45 seconds. Exposing the 
100 
90 
80 
70 
60 
50 
40 
50 
20 
10 
DURATION OF AFTER-1 MAGE ( SECONDS) 
■S=- 
_\<_P 
V? 
50° 
■20p- 
100 
0 
1 
2 
3 
4 
EXPOSURE (SECONDS) 
Fig. 34. Showing the duration of after-images from exposures of 
various lengths to light-sources of different brightnesses. 
eye to it for 3.5 seconds increased the duration of the after- 
image to 93 seconds. As is well known, the after-image 
is blinding at first and, when it is considered that vision is 
interfered with for an appreciable period, it is not surprising 
that production increases in offices and factories when such 
exposed sources are replaced with suitable shades or 
diffusing equipment. 
The frequency at which a flicker disappears depends 
upon the maximum brightness and upon the wave-form. 
The vanishing-flicker frequency increases with the bright- 
ness and is greater for a cycle in which the change in  LIGHT AND WORK 
170 
brightness is abrupt than for one in which the change is 
gradual.36 Owing to the lag in cooling and heating of a 
lamp filament we generally have a flicker superposed on a 
steady average brightness and it is invisible excepting at low 
frequencies. For 25-cycle electric energy this superposed 
flicker is apparent on surfaces that are relatively bright. 
For example, a certain intensity of illumination may cause 
a surface of 80 per cent reflection-factor to be bright enough 
so that its brightness variation is noticeable. The same 
intensity of illumination on a dark gray surface of 8 per 
cent reflection-factor, would cause a brightness of only one 
tenth of that in the former case. This brightness might 
be too low in value to appear to vary under the 25-cycle 
electric energy. The matter of flicker is touched upon here 
chiefly to emphasize this phase of vision. However, it is of 
practical interest in lighting on low-frequency electric 
energy and in various special fields such as the motion- 
picture screen. 
It is beyond the scope of this book to discuss this phase 
of vision in detail but it appears worth while to present the 
data in Fig. 35 in order to give the reader, who may be un- 
familiar with the subject, an idea of the frequencies at 
which flicker vanishes. In this case the cycle of varying 
brightness is one-half bright and one-half dark and the 
change from one to the other is abrupt. It is produced by 
means of a rotating disk, with equal open and opaque 
spaces, placed between a light-source of small dimensions 
and the surface illuminated. The frequency of flicker 
(cycles per second) is increased until the flickering bright- 
ness of the illuminated surface disappears. Inasmuch as 
the brightness of the surface depends on both the intensity 
of illumination and the reflection-factor of the surface, 
curves are plotted of vanishing-flicker frequency for sur- 
faces of different reflection-factors for the same range of 
illumination in foot-candles. It is seen that for surfaces FUNDAMENTALS OF VISION 
171 
in general the rise in this critical frequency is most rapid 
below five foot-candles. It should be particularly noted that 
this is a case of complete and abrupt change from bright 
to dark and that the period of bright and dark are equal. 
The author 30 has shown the effect of wave-form of the 
60 
50 
40 
30 
20 
10 
0 
6O'I» Reflection-factor 
40'1' 
zo*/. 
10°/- 
5'1* 
VANISHING-FLICKER FREQUENCY 
(CYCLES per second) 
5 
10 
15 
20 
FOOT-CANDLES 
Fig. 35. The frequency at which flicker disappears for surfaces of various 
reflection-factors illuminated to various intensities as indicated by the horizon- 
tal scale. The flicker was caused by abrupt changes from bright to dark, 
the periods of bright and dark being equal. 
cycle elsewhere, but the oniy cases where flicker is more per- 
sistent is when the period of darkness is much greater than 
the period of light. 
External Factors in Vision 
Although in previous chapters many of the fundamen- 
tals of vision have been presented, there are many other 
details of vision that have not been touched upon. In 
general, we see by virtue of our ability to distinguish dif- 
ferences in brightness and in color. The eyes and the visual 
processes involved are not the only factors in vision. There 
are external factors equally important, such as the direction, 
the diffusion, the quantity, the distribution, and the quality 172 
LIGHT AND WORK 
of light. Shadows play a part as well as high-lights. Back- 
grounds are sometimes very important. In other words, 
if we are to produce the best conditions for seeing, we 
must aid the visual sense by giving the best conditions to 
these external factors. If we had only the eyes to consider 
there would be little that we could do to aid vision except- 
ing to furnish the necessary light. The data presented in 
this chapter are generally for two eyes but the test-objects 
were two-dimensional; that is, they were not in relief. For 
this reason some of the advantages of binocular vision over 
monocular vision have not been touched upon. 
Binocular Vision. Two eyes have a great advantage 
over one eye when objects in relief are to be distinguished. 
Each eye sees a slightly different aspect of a three-dimen- 
sional object and this plays a great part in giving the effect 
of relief. Incidentally, one of the most marvelous features 
of vision is that we see objects out in space where they are 
instead of seeing them " internally " where the retinal 
images are or in our brain. This is one of the greatest 
examples of learning. The various muscular stresses such 
as those due to convergence of the eyes, we have learned to 
interpret in terms of external space. Two eyes greatly 
extend our interpretation of the external world.42 Knowl- 
edge plays an important part and where our knowledge is 
supposedly accurate, but is actually erroneous, we misjudge 
or are confronted with illusions. In any thorough study 
of lighting conditions in relation to vision, the added advan- 
tages of binocular vision should be taken into consideration. 
The principal modes by which we perceive the third 
dimension of space and of objects and other aspects of the 
external world are as follows: 
a. Extent 
b. Perspective 
c. Elevation of objects 
d. Variation of light and shade on objects FUNDAMENTALS OF VISION 
173 
e. Variation of visor angle in proportion to distance 
j. Muscular effort attending accommodation of the 
eyes 
g. Muscular effort attending convergence of the axes 
of the eyes 
h. Interference of near objects with those more distant 
i. Stereoscopic vision 
j. Clearness of brightness or color as affected by dis- 
tance (of interest only outdoors for great 
distances) 
Shadows. The area on an object which does not receive 
the dominant direct light we will term the shadow. The 
shadow of an object cast upon another surface we will term 
the cast shadow. These play very important parts in seeing 
objects.6 Without them objects would be invisible when 
the background was of the same brightness and color. 
Sometimes it is the shadow cast by the object upon another 
surface that reveals the object to us. The size of the light- 
source or, to be exact, the solid angle subtended by the 
light-source determines the character of the shadow-edge. 
The sun causes a fairly sharp shadow; the sky a very in- 
definite one. So it is with artificial lighting equipment. 
Small sources of light cause sharp shadows. Larger diffusing 
glass units produce shadows with softer edges. Ceilings 
produce effects corresponding to the sky. The position of 
the light-source determines the position or direction of the 
shadow. The amount of scattered light from the ceiling 
and other areas determines the brightness of the shadow. 
If details are to be seen in shadows the latter should be 
luminous. In such cases the shadow should receive not less 
than io per cent as much light as an adjacent surface 
under full illumination. Of course, this cannot be adhered 
to strictly but it represents at least a limit of desirability. 
When details are not to be seen in the shadows, vision will  LIGHT AND WORK 
174 
usually be aided if the shadows are very dark. In general, 
multiple shadows are confusing to vision, so that it is de- 
sirable to have a predominant light-source. 
Highlights. These are the complementaries of shadows 
and they are important much in the same way. The object 
is often distinguished by the contrast between highlight 
and shadow and the form is often distinguished by the 
character of the highlight or by the modulation between 
highlight and shadow. A cylindrical polished object has a 
linear highlight, a spherical one a small highlight, etc. 
Sometimes we distinguish an object only by its highlight. 
This is particularly true of small objects, such as fine wire. 
Backgrounds. Inasmuch as contrast is essential to vision 
our control of background can be used to great advantage. 
We can make the background very much brighter or very 
much darker than the object we wish to see. The smaller 
the object the greater is the contrast necessary to see it. 
For opaque objects we can use illuminated backgrounds 
of as high brightness as necessary. This is a very satisfac- 
tory method of seeing in many cases, such as photographic 
negatives. However, it is often impracticable and is gen- 
erally more trying on the eyes than the dark background. 
A fine polished wire can be seen best as a brilliant highlight 
on a dark background. The latter can be obtained prac- 
tically black by providing a hole in a box lined with black 
velvet. Practically no light comes out of this hole, there- 
fore it is nearly perfectly black. The highlight on the fine 
wire can be easily obtained by means of a local light-source 
well shaded from the worker's eyes. This kind of fine work 
requires local light-sources supplemented by general lighting. 
A few years ago there was a tendency to condemn localized 
light and to solve all lighting problems by general lighting. 
It was obvious at that time that those advocating such a 
procedure were not thoroughly familiar with the details 
of lighting and vision. FUNDAMENTALS OF VISION 
175 
Color can be used in the background sometimes, but it 
should be a shade of a tint of a mid-spectrum hue, such as 
green, yellow-green, or yellow. In many places one will 
find operators working without plain backgrounds against 
which to see the work notwithstanding the installation of 
backgrounds would be simple, inexpensive, and effective. 
In the use of bright backgrounds against which dark 
details are seen, it should be noted that irradiation - the 
spreading of the bright areas over the dark areas - tends 
to decrease the apparent size and therefore the visibility 
of the dark details. On the other hand the fine detail seen 
as a highlight is apparently enlarged and therefore more 
visible. An exaggerated case is a glowing filament of a 
lamp. It appears many times larger than it actually is. 
Even at a low voltage when it barely glows it appears much 
larger than it is. When the filament is " cold " it is 
scarcely visible. 
An interesting experiment of the effect of background 
can be performed by stretching fine wires across strips of 
white, gray, black paper and black velvet. By experiment- 
ing with highlights on the wires, and by using dark and light 
wires, some interesting observations can be made. Chapter IX 
SPEED OF VISION 
SEEING is a commonplace achievement in which we are 
seldom confronted with the element of time. Even more 
rarely are we conscious of time as a factor. However, work 
is an integral in which time is a vital factor, for the output 
of any worker or group of workers is directly influenced by 
the " speed " at which individual acts are accomplished. 
In general, any operation at which a worker is engaged 
involves visual discrimination, mental decision, and mus- 
cular activity. The relative importance of these may vary 
considerably with different operations, but vision in most 
cases plays an important part. The trained operator be- 
comes almost an automaton, accomplishing the various 
visual, mental, and muscular acts almost unconsciously or 
at least in apparent abstraction. 
Workers may differ in output for a given time because of 
the fundamental individual difference in alertness or speed 
of reaction. In all operations involving vision, and all do 
to some degree, the individual must see what it is necessary 
to see. He may do so quite unconsciously, nevertheless he 
must discriminate. For example, if he is inspecting material 
for flaws, he must see the flaws or overlook them. If they 
are present and he does not see them, he has not done his 
work well. If the lighting is inadequate or improper, his 
ability to see is reduced and he must look slowly and care- 
fully and perhaps go over the same ground more than once. 
This consumes time and if the visual part of the entire 
operation is appreciable, the time consumed is appreciable. 
Of the time consumed in visual discrimination, a portion of 
it is lost unless the lighting is best for that particular prob- 
176 SPEED OF VISION 
177 
lem of seeing. This would be very obvious if the visual 
act required a long time, like looking for a fine mark on an 
object. However, it is just as true in operations in which 
each individual visual act occupies only a short time. It 
should be obvious that lighting has a great influence on 
ease of seeing; therefore, we would expect lighting to play 
an important part in the output of most workers. As a 
matter of fact it does and to an extent unsuspected by most 
persons. One needs only to observe persons at work here 
and there to be convinced of the handicaps of improper 
and inadequate lighting. The eyes may be squinted, the 
head screwed around, the object turned over and what not 
before the visual act is completed when it could have been 
achieved under proper and adequate lighting in a straight- 
forward manner with less strain on the operator and lower 
cost. 
The speed of discrimination or speed of vision, as it is 
termed for the sake of brevity, involves all factors of light 
and lighting which influence our ability to see differences in 
brightness, color and fine detail. In the general case those 
factors influencing the discrimination of color also may be 
important. However, in actual practice after the quality of 
light is chosen, the chief factors of light and lighting are 
the intensity of illumination, the distribution of light and 
brightness, and the direction of light. Of these, the first is 
still of interest even after the others have been taken care 
of in proper lighting, for we have the possibility of increas- 
ing the intensity of illumination to a very high value before 
reaching a point where the advantage to vision ceases to 
increase correspondingly. This point is well proved in Chap- 
ter VIII. Some of these visual laws have been known for 
years. Furthermore, reaction-time has been studied for 
years and it has long been known that intensity of illumina- 
tion and brightness are important factors in the time 
involved in the act of discrimination. However, for some 178 
LIGHT AND WORK 
reason (perhaps the previous relatively low efficiency of 
light production) the time element in seeing has not received 
much consideration until recently. 
Tests in the industries are showing that production is 
increased when better lighting is substituted for poor light- 
ing. Some of the data are presented in Chapter X. Valu- 
able as these tests are, it is necessary to depend upon 
laboratory experiments to analyze the possibilities. Ex- 
periments in the industries should be continued, as doubtless 
they will be, but these are not subject to the control neces- 
sary to unravel the complexity of human activity which 
involves vision. Thus we need and welcome the data 
obtained in laboratory researches. 
The effects of quality or spectral character of light have 
been discussed in other chapters; therefore, for the present, 
only the effect of intensity of illumination will be considered. 
Unless otherwise stated, all other factors are maintained 
constant in each of the researches described herewith. 
The illuminant in all cases is the light from tungsten 
filament lamps which has a continuous spectrum as shown 
in Fig. 19. Its color is a yellowish white; that is, a very 
unsaturated yellow. 
Inasmuch as most laboratory tests are made with black 
characters on a white background, it should be noted that 
this is a much better condition of visibility than the lower 
contrasts involved in most industrial operations. For this 
reason, it is well to perform laboratory tests with objects 
of reduced contrast, such as black on medium gray or dark 
gray. In general, it is found that visibility and speed of 
discrimination are of lower values under a certain inten- 
sity of illumination when the test-object is of low-contrast 
than when it is black and white. 
Speed of Discrimination of Very Fine Detail 
Although most of us seldom find it necessary to dis- 
criminate the finest detail that we can see under a given SPEED OF VISION 
179 
condition of lighting and surroundings, this so-called visual 
acuity is an important phase of vision from a scientific 
viewpoint at least. The data yielded by tests of visual 
acuity are also of significance in everyday work, but they 
are often overestimated. When applied to any particular 
eye they are usually a measure of the " factor of safety " 
of the eye in the discrimination of detail because most eyes 
are not required to do much discriminating of details at 
the threshold of visibility. In Chapter VIII it is seen that 
visual acuity increases as the intensity of illumination in- 
creases. The time required to discriminate or to recog- 
nize (that is, to see) a small object, decreases as the in- 
tensity of illumination increases. The speed of discrimina- 
tion is the reciprocal of this time usually in seconds. The 
size of test-objects used in visual-acuity tests is given in 
terms of visual angle because such a specification involves 
both the actual size of the object and its distance from the 
eye. For objects very small in comparison with their dis- 
tance from the eye, the visual angle is approximately pro- 
portional to size of the object divided by its distance. 
In order to illustrate the effect of intensity of illumination 
on the speed of discrimination of fine details, the work 
of Ferree and Rand 49 will first be drawn upon. By speed of 
discrimination is meant the reciprocal of the time in seconds 
required for the observer to see the missing portion of the 
split ring. They used the international test-object con- 
sisting of a black ring on a white ground, a portion of the 
ring being missing. In other words, it is a black split ring 
on a white ground as shown in Fig. 36. The visual angle 
is that subtended by the dimension of the missing portion a 
of the black ring at the distance of the eye from it. This 
ring is rotated into various positions so that the observer 
must not only distinguish the missing portion, but, to make 
certain that he does, he is obliged to state in what position 
this missing segment is. 
Ferree and Rand employed four visual angles, 1.15, 1.73, 180 
LIGHT AND WORK 
2.49, and 3.45 minutes of arc, and illuminated the test- 
object with unaltered light from tungsten filament lamps. 
The visual angle of 2.49 minutes of arc is approximately 
equivalent to the visual angle subtended by the details of 
10-point type at a distance of 13 inches. In order to give 
a further idea of small visual angles some data have been 
presented in Table XI. These data are plotted in Fig. 25. 
a 
Fig. 36. The international test-object for visual acuity. The 
dimension of the opening a determines the visual angle when distance 
from the eye is taken into account. The subject not only must see 
the discontinuity of the ring but must state in what position it is. 
Several sizes and positions are reproduced herewith so that the 
reader may compare them at a distance at which the smallest is not 
seen as a split ring. 
The smallness of detail which can be distinguished by 
normal eyes depends largely upon retinal structure, but, for 
defective eyes, astigmatism and other errors of refraction 
may play a major part in lowering visual acuity. The 
limiting visual angle for normal eyes is about 40 seconds or 
0.67 minutes of arc. That is, two objects separated by less 
than 0.67 minutes visual angle cannot be seen as being 
separated by most eyes that can be classed as normal. 
In Fig. 37 are plotted the relation between intensity of 
illumination and the speed of discrimination of the inter- 
national test-object (black split ring on white paper of 
80 per cent reflection-factor) of the four different sizes or 
visual angles as obtained by Ferree and Rand. It is seen 
that over a range of intensity of illumination up to 12 foot- 
candles for the black-white test-object, the speed of dis- 
crimination increases and shows signs of continuing to 
increase far beyond this highest intensity of illumination. SPEED OF VISION 
181 
If the test-object was truly black and the background was 
a dark gray of 8 per cent reflection-factor, the identical 
results shown in Fig. 37 would have been obtained under 
illumination intensities exactly ten times the values shown 
in the lower horizontal scale. For this reason the upper 
FOOT- CANDLES 
(8$ reflection-factor) 
240 10 40 60 80 100 IZO 
ZOO 
160 
IZO 
80 
40 
SPEED OF DISCRIMINATION 
( RECIPROCAL OF TIME INSECONDi) 
1.15 
1.15* 
O 2 4- 6 8 IO 12 
FOOT-CANDLES 
(60*l« REFLECTION-FACTOR) 
Fig- 37- Showing the speed of discrimination as affected by 
intensity of illumination (brightness of background) for black 
test-objects of four different sizes (visual angles 1.15, 1.73, 2.49 
and 3.45 minutes. The lower scale represents the illumination of 
a colorless background of 80 per cent reflection-factor against 
which the test-object was viewed. If the background were a 
dark gray of 8 per cent reflection-factor, the foot-candles must be 
multiplied by 10. Thus the upper scale fits the data equally 
well assuming that the test-object was truly black on the back- 
ground of 8 per cent reflection-factor. 
scale is applied. If the upper and lower scales are reduced 
to brightness scales they will be identical. If any other 
reflection-factor of background is to be considered, the 
foot-candle scale can be constructed accordingly. It is 
only necessary to reduce it to the same brightnesses as the 
two scales given. Another point should be mentioned and 
that is, the fact that the so-called black of the test-object 182 
LIGHT AND WORK 
was not truly black. It reflected a slight amount of light - 
maybe as much as 2 or 3 per cent. Assuming it to be on 
the dark gray background of 8 per cent reflection-factor, 
it is seen that the contrast would be considerably less than 
true black on this same dark gray. However, this would 
require even greater intensities of illumination than those 
of the upper scale for the same speeds of discrimination. 
In Fig. 38 the per cent increase in the speed of discrimi- 
nation is plotted against foot-candles for the smallest and 
FOOT-CANDLES 
(S'l- REFLECTION-FACTOR) 
600 
ZO 
40 
60 
80 
100 
IZO 
PER CENT INCREASE IN SPEED OF DISCRIMINATION 
500 
A, 
400 
300 
zoo 
100 
0 Z 4 6 8 10 12 
FOOT-CAN DLE5 
(80'Io REFLECTION -FACTOR) 
Fig. 38. Data in Fig. 37 for the smallest and largest test- 
objects are plotted in terms of per cent increase in speed of 
discrimination for the same range of foot-candles on the two 
backgrounds (so-called white and dark gray). 
the largest test-objects (visual-angles) represented in Fig. 
37. It is seen that the speed of discrimination increases 
more for the smallest test-object. The curves for the two 
intermediate test-objects if plotted from Fig. 37 on Fig. 38 
would fall between those for the largest and smallest. It 
is seen that the per cent increase in speed of discrimination 
is greater for the smaller test-object. SPEED OF VISION 
183 
In this investigation 13 observers were used. It is to be 
expected that the effect of increase of illumination becomes 
less as the visual angle or size of the test-object is increased; 
however, there is good reason for believing that there is 
a considerable effect for objects subtending much larger 
visual angles. Certainly for the lesser contrast (upper 
scale) speeds of discrimination, comparable with those of 
black-white contrasts of the greater values, require illumi- 
nation intensities of 100 foot-candles or more. 
240 
ZOO 
160 
120 
60 
40 
O I Z 5 
SPEED OF DISCRIMINATION 
(reciprocal OF TIME IN seconds) 
- 
- 
SIZE OF TEST-OBJECT IN MINUTES OF VISUAL ANGLE 
Fig. 39. Showing the relation between visual angle and speed of 
discrimination for three intensities of illumination on the black-white 
test-object. If the test-object were truly black and were seen against 
a background of about 8 per cent reflection-factor these foot-candle 
values would be ten times greater. 
The relation between visual angle and speed of discrimi- 
nation for three intensities of illumination on the black- 
white test-object is shown in Fig. 39. In plotting these 
data the author has taken the liberty of drawing straight 
lines because they " fit " the data fairly well and simplify 
the diagram somewhat. It is seen that the time required 
to recognize the test-object decreased (speed of discrimina- 
tion increased) as the visual angle increased. Furthermore, 184 
LIGHT AND WORK 
the increase of discrimination as the visual angle increased 
was greater for a high intensity of illumination than for a 
low one. 
In general, these data show that speed of vision (of see- 
ing, of recognition, of discrimination) increases as the 
intensity of illumination increases. By extension it may be 
safely predicted that, other conditions being equal, a group 
of workers would turn out more work under a high intensity 
of illumination than under a relatively low one, when the 
work involved discrimination of fine details such as in weav- 
ing, fine machine work, close inspection, and many other 
factory operations. The increase in output due to increased 
intensity of illumination doubtless would depend, to a de- 
gree at least, upon the relative proportion of the visual 
discrimination to the other parts of the complete operation. 
In all cases it is seen that the advantage of increasing 
illumination is very marked between 0.4 foot-candles and 
12 foot-candles (for black details seen against a so-called 
white background) and that it promises to continue far 
beyond 12 foot-candles. Finally, it should be recalled that 
this test-object was " black " on " white." For lesser con- 
trasts the foot-candle scale must be multiplied by a factor 
greater than unity; that is, the percentages of increase for 
intensities in the region from 2 foot-candles to 10 foot- 
candles would be greater even than with the high-contrast 
test-object which was used. Much of our vision is con- 
cerned with lesser contrasts and the foot-candles required 
increase directly as the reflection-factor of the background 
decreases for black details seen against it. For other lesser 
contrasts, for example, light gray on white or dark gray on 
medium gray, the intensities required for equal visibility 
and speed of discrimination are much greater than for black 
on white. Therefore, the upper foot-candle scale has been 
introduced in Figs. 37, 38 and 39. (See Figs. 56 and 57.) 
Dr. P. W. Cobb,45 who for many years has contributed SPEED OF VISION 
185 
results of extensive researches carefully executed, has re- 
cently begun to report results of various experiments on 
the time required to recognize an object. Much of this 
work is particularly important in analyzing the various in- 
fluences which aid or hamper vision. He used several 
test-objects. Some of them appeared to the observer out 
of a clear field; others were preceded and followed by 
" confusion patterns." In the former cases, the observer 
fixed his eyes on a point where the dot was to appear but 
the field was without pattern. In the latter cases, the 
observer fixed his eyes upon the point where the test-object 
was to appear, but before it appeared another pattern was 
presented for a short interval of time. This was followed 
immediately by the test-object and it in turn was followed 
by another pattern. 
In Fig. 40 is shown the effect of illumination intensity 
on the speed of discrimination of a circular black dot on a 
so-called white background of 80 per cent reflection-factor. 
The black dot was 0.167 inches in diameter and was viewed 
at a distance of 19.7 feet. Its diameter, therefore, subtended 
an angle of 2.43 minutes at the subject's eyes. The dot 
was actually black; that is, its brightness was zero. 
The dot was preceded and followed by a blank field of the 
same brightness as the background upon which the dot 
was seen. In such experiments it is essential to know 
whether or not the observer really sees what he says he 
does. In these experiments the exposures were made in 
groups of ten consecutive ones. The dot was present five 
times and absent five times, the order of absence and pres- 
ence being haphazard. The observers reported when the 
dot was present and when it was not. In each subsequent 
group of ten, the exposure was shortened until the results 
showed that the dot was no longer seen with certainty. 
It is seen in Fig. 40 that the speed of discrimination is still 
increasing at 130 foot-candles (on the white background) 186 
LIGHT AND WORK 
and that at illumination intensities below io foot-candles, 
which are those most prevalent indoors, the speed of seeing 
the black dot is far below its value at 50 foot-candles, for 
example. It is to be noted that the maximum speed has not 
been reached at 130 foot-candles on the white background 
FOOT-CAN DLE5 
(8'l' REFLECTION-FACTOR) 
150 
140 
130 
120 
no 
100 
90 
60 
70 
60 
50 
40 
30 
ZO 
10 
ZOO 
400 
600 
800 
1000 
1200 
1400 
5PEED OF DISCRIMINATION 
(reciprocal of time) 
0 10 ZO 30 40 50 60 70 60 90 100 110 IZO 130 140 150 
FOOT-CANDLES 
(a0°|o REFLECTION-FACTOR) 
Fig. 40. Showing the speed of discrimination of a black dot (2.43 minutes 
visual angle) on a so-called white background (80 per cent reflection-factor 
for a range of illumination up to 130 foot-candles. The upper scale is for a 
truly black dot on a dark gray background of 8 per cent reflection-factor. 
and there is no indication that it will be reached until 
much higher intensities are reached. This is in line with 
much of the data pertaining to visual laws which have been 
available for a long time. Furthermore, it is what would SPEED OF VISION 
187 
be expected from eyes and visual processes adapted for 
countless thousands of years to very high intensities 
outdoors. 
Inasmuch as this test was for a black dot on a white back- 
ground, it is of interest to inquire what would be the cor- 
responding intensities for the same black dot when seen 
on a background of low reflection-factor. As has been 
done in the previous cases, a very dark gray background 
of 8 per cent reflection-factor is taken as representing a 
common extremely low contrast. The corresponding foot- 
candle scale is added at the top of Fig. 40. It is now 
seen that a point just above the bend of the curve cor- 
responds to about 500 foot-candles for this low-contrast 
test-object; however, as Dr. Cobb has pointed out, the 
position of the knee of the curve depends upon the range 
of the scale plotted. This is discussed in Chapter XIII. 
Speed of discrimination is still increasing even at 1000 
foot-candles. It is recognized in this case the dot was truly 
black, so that this transformation from lower to upper scale, 
in replacing the background of 80 per cent reflection-factor 
with one of 8 per cent reflection-factor, is strictly correct. 
Usually " black," reflecting some light as it does, results 
in a lower contrast than the low one represented by the 
upper scale. Hence, the foot-candle values in the upper 
scale should be even larger in most cases. 
Speed of Reading 
The choice of a test-object for a visual test is always a 
matter of serious consideration. It should be one easily 
described and reproduced; it should be such as to yield 
results that may be interpreted to some extent in terms of 
everyday visual operations; and it should be one that is 
sensitive enough to provide a measure of the influences 
under consideration. Reading is a very commonplace ac- 
tivity and reading matter can be accurately described and LIGHT AND WORK 
188 
reproduced. Desirous of obtaining some measure of the 
effect of illumination intensity upon the speed of reading, 
the author and his colleagues chose reading matter in one 
of the initial investigations.22 
Contrary to general opinion, ordinary reading is not 
solely a matter of visual acuity or the ability to see fine 
detail. Visual acuity is a measure of the finest detail that 
can be seen under a given set of conditions. The individual 
letters of a printed page are rarely so small as to be just 
visible under ordinary lighting conditions. In fact, they 
are many times larger than the smallest that we could see 
under the conditions. Furthermore, it is known that in 
reading, the eyes do not move at a uniform speed across 
the page. They jump from point to point, pausing in steady 
fixation usually from two to seven times across the ordinary 
four- or five-inch line. During these jumps the eyes cannot 
possibly see clearly any more than they could see rapidly 
moving print when they were fixed upon a stationary point. 
Clear vision not only takes place when the eyes are sta- 
tionary, but, as shown in Fig. 28, visual acuity is marked 
only at the point of the retina intersected by the optical 
axis of the eye. In other words, ordinary reading involves 
recognition of groups of letters, and even of words, a great 
deal larger than the smallest which can possibly be seen 
under the average conditions of lighting. For this reason 
the results of visual acuity tests are not necessarily the 
same as those of reading and most visual acts, although 
there is bound to be a similarity. 
A further reason for adopting the reading test was to 
develop a method of measuring the effect of illumination 
intensity on speed of vision without requiring a summary 
of data; in other words, to devise a " direct-reading " ap- 
paratus. A number of test-objects were used, such as lines 
of various kinds, letters and other characters, and different 
styles of reading matter. The contrast in the test-objects SPEED OF VISION 
was chiefly that resulting from " black " ink on " white " 
paper, but later this contrast was reduced in the case of the 
most successful reading test-object (reading matter in Old 
English type) by decreasing the reflection-factor of the 
paper. The number of observers varied from 37 to 49 
persons under middle age. 
Eventually, after using various test-objects, the general 
plan adopted was to have an observer read aloud from a 
column of reading matter attached to a revolving drum C, 
Fig. 41, the speed of which he adjusted to correspond to his 
189 
A 
K 
H 
C 
C 
£ 
F 
Fig. 41. Apparatus used for speed of reading tests. 
Old English print revolved on the drum C under the 
slit E which was viewed from H with the head resting 
on K. The light-source was G in the shield F and the 
intensity was controlled without change in quality. 
D was a screen of the same general brightness of the 
printed matter at all times. 
B 
highest rate of reading under a given intensity of illumina- 
tion. A screen D with a horizontal slot E at the center con- 
cealed from view everything excepting three or four lines 
of the Old English reading matter. The screen was of the 
same general brightness as the white page at all times. 
When the speed of the drum and the observer's highest 
rate of reading did not agree, his reading would be inter- 
cepted by either one or the other edge of the slot. The 
speed of the drum, when properly adjusted by the reader  LIGHT AND WORK 
for his maximum speed of reading under the particular 
illumination intensity, was read from the miles-per-hour 
scale of an automobile speedometer which was attached to 
the motor which drove the drum. The speed of the motor 
was controlled by means of a rheostat. The observer placed 
his forehead at K and looked through the slot H. The light- 
source was confined in a diffusing glass cylinder G and illu- 
minated the reading matter through a slot in the opaque 
cylinder F. In order not to vary the color of the light, G 
was covered with tinfoil from which pointed " teeth " were 
cut. These teeth moved across the slot in F as G was turned 
and the percentage of the opaque area varied from zero 
to ioo for one revolution of the drum. 
The Old English type was finally chosen because the rate 
of reading it was more uniform inasmuch as the reader was 
unable to recognize any word or group of letters much more 
readily than any other. He had to give strict attention to 
reading this less familiar type. He was required to read 
aloud because this required both mental and physical ac- 
tivity and made certain that he was actually reading every 
word. The type was clear-cut and uniform and was in the 
form of two narrow columns without paragraph inden- 
tions. It contained no uncommon words and the subject 
matter was such as would not require unusual thought or 
concentration on the part of the reader. Incidentally, 
such a test also introduces a condition not usually found 
in laboratory tests, but which is almost always found in 
practice. That is, the retina is pre-exposed all the time to 
various images so that the effect of after-images of various 
patterns is always present as it usually is. Other details 
will be found in the original paper?2 
The results of this investigation are plotted in Fig. 42, 
for reading black print on white paper (Curve B) and black 
print on gray paper (Curve A). The reflection-factor of 
the white paper was 80 per cent and of the gray paper 
190 SPEED OF VISION 
was about 23 per cent. On increasing the intensity 
of illumination from 0.4 foot-candles to 4 foot-candles, 
the speed of reading increased 54 per cent for the 
high-contrast test-object. On increasing the intensity of 
illumination from 4 foot-candles to 16 foot-candles the 
speed of the reading increased 15 per cent for the high- 
191 
160 
150 
140 
130 
120 
no 
IOG 
■90 
* 
70 
A, 
Speed of Reading, Percent. 
B 
0 4 8 I? 16 20 24 28 
Intensity of Illumination. Foot~Candles 
Fig. 42. Effect of intensity of illumination on the speed of 
reading. A is for a low-contrast test-object (black print on 
a background of 23 per cent reflection-factor). B is for the usual 
condition of black print on a white paper of 80 per cent reflec- 
tion-factor. 
contrast test-object and 50 per cent for the low-contrast 
test-object. It is seen that the influence of intensity of illu- 
mination is very marked in both cases but much more so 
for the low-contrast test-object. The latter contrast more 
nearly approaches the contrasts ordinarily encountered in 
vision than the black-white contrast does, with the excep- 
tion, of course, of reading and some other work where 
white backgrounds are employed. It will be noted that 192 
LIGHT AND WORK 
the two curves are made equal at 4 foot-candles. Of course, 
the speed of reading A was actually much slower than B. 
It is also of interest to plot these data in terms of the 
brightness of the background as in Fig. 43 instead of in 
terms of foot-candles as in Fig. 42. Now the results would 
be the same if the black ink was really black or non- 
reflecting. But ink is far from black and the results still 
140 
130 
120 
no 
100 
90 
80 
100 I 2 3 4 5 6 7 8 9 10 
4 
Speed of Reading, Percent 
B 
Brightness of Background. Millilomberts. 
Fig. 43. Relation of brightness of background (the blank page) and 
speed of reading. (Data from Fig. 42.) A is for low contrast (black 
print on gray paper of 23 per cent reflection-factor) and B is for high 
contrast (black print on white paper of 80 per cent reflection-factor). 
differ as shown in Fig. 43. This is of particular interest in 
indicating that the foot-candle scales (for surfaces of 8 per 
cent reflection-factor), which have been inserted on a num- 
ber of illustrations in this and the preceding chapter, are 
above criticism. That is, the values should really be in- 
creased, as intimated several times, to allow for the effect 
of reflection of light by the so-called black test-objects in 
decreasing the contrast to a lower value than that due to a 
perfect black on a dark gray of 8 per cent reflection-factor. 
(See Figs. 56 and 57.) 
It is seen in both Fig. 42 and Fig. 43 that the speed of SPEED OF VISION 
reading shows signs of still increasing beyond the highest 
intensity of illumination (30 foot-candles) showing that 
the advantage of increasing illumination extends beyond 
this. This indication is very marked in the case of the low- 
contrast reading matter. 
193 
Speed of Discrimination as Affected by " Confusion- 
Patterns " 
Let us again take up some of the more recent work of 
Dr. P. W. Cobb 45 in which he introduces the confusion field 
before and after the exposure of the actual test-object. His 
test-object consisted of two black bars as shown in b, Fig. 
44. These could be rotated so as to present them in various 
a 
b 
c 
Fig. 44. The test-object b is preceded by a and followed by c. These 
confusion fields a and c have a marked influence upon the time required to 
distinguish b. 
positions to the observer. The two other patterns, a and c, 
are confusion fields introduced respectively before and after 
the test-object b was presented. In Fig. 45 the curve A 
shows the relation of speed of discrimination of the test- 
object b preceded and followed by a blank field of the same 
brightness as the background of the test-object b. The 
curve B represents the relation found between speed of 
discrimination and intensity of illumination of the same test- 
object, but in this case the exposure of the test-object was 
preceded and followed by the confusion patterns a and c  LIGHT AND WORK 
194 
respectively, Fig. 44. The test-object consisted of two 
black bars, each 0.125 inches in width, and they were sepa- 
rated by a white interval of equal width. The length of 
the bars was three times the width of each bar so that the 
two black bars and the intervening white interval covered 
FOOT-CANDLES 
(s°|o reflection-factor) 
zoo 
160 
160 
140 
120 
100 
80 
60 
40 
20 
0 10 20 ~ 30 4 0 50 60 70 80 90 100 110 
200 400 600 800 <000 
A 
SPEED or DISCRIMINATION 
(RECIPROCAL OF TIME) 
B 
F00T-CANDLE5 
REFLECTION-FACTOR) 
Fig. 45. Curve A is the relation of the speed of discrim- 
ination of the test-object b, in Fig. 44, to the intensity of 
illumination of the white background of 80 per cent reflec- 
tion-factor. Curve B shows the results when the test- 
object b is preceded by a and followed by c, Fig. 44. The 
upper scale represents the relation between speed of dis- 
crimination and foot-candles if the test-object had been 
truly black on a dark gray background of 8 per cent 
reflection-factor. 
a square area. In other words, this test-object could be 
produced by removing a middle one-third from a black 
square 0.375 inches on a side. The visual angle subtended SPEED OF VISION 
195 
by 0.125 inches at the distance of 19.7 feet employed in 
these experiments is 1.82 minutes of arc. The reflection- 
factor of the " black " bars was about 4 per cent and of the 
white background 80 per cent. Curves A and B represent the 
average results of four observers and a long series of care- 
ful tests with approved apparatus and methods. Subse- 
quent tests with other observers gave similar results. 
It is seen on comparing these two curves that the intro- 
duction of the confusion element, due to patterns a and c, 
Fig. 44, has materially decreased the speed of discrimina- 
tion. It was to be expected that the variegated images 
continually falling in succession upon the retina in the 
course of everyday vision constitute a confusion to some 
degree at least, interfering with the seeing of any particular 
pattern of the series. These interfering after-images are not 
only visual but are also mental to some extent. It is known 
that after-images are more persistent as the brightness of 
the object increases, and Dr. Cobb's work indicates that 
this confusion is more effective at higher brightnesses than 
at relatively lower ones. 
Dr. Cobb also worked with a larger test-object of the 
same form preceded and followed by confusion field cor- 
respondingly enlarged. In this case the test-object and 
confusion patterns were drawn to double the scale of the 
test-object used in connection with curves A and B. That 
is, the visual angle was that subtended by the distance (0.25 
inch) between the black bars a, Fig. 44, at the distance of 
19.7 feet. This visual angle was 3.63 minutes. For the 
larger test-object the speed of discrimination was greater 
than that of the smaller test-object for the same intensity 
of illumination. If plotted, the curve would be about 50 
per cent higher than curve B above the " knee " of the 
latter. An idea of the angular size of these test-objects in 
terms of the same angular size of objects at a distance of 
14 inches can be gained from Fig. 25. In all these experi-  LIGHT AND WORK 
196 
merits the results show the advantage of higher intensities 
of illumination and that the advantage is still apparent 
above 100 foot-candles on the white background and above 
iooo foot-candles if the test-object had been a true 
black on a dark gray background of 8 per cent reflection- 
factor. 
Simple Demonstrations 
The increase in " speed of vision " with increase of illu- 
mination intensity can be demonstrated in various ways by 
anyone. An observer can be asked to read aloud a certain 
amount of printed matter under a low intensity of illumina- 
tion and the time consumed can be recorded. He can then 
read the same amount of material under a much higher 
intensity. The time in the latter case will be found to be 
shorter than in the former case. It is better not to have 
the reader read the same material in the two cases. 
Another simple way of demonstrating this is by exposing 
such a letter as E for different periods of time under dif- 
ferent intensities of illumination. A slot may be cut in a 
white cardboard and a letter E pasted on another cardboard. 
This E is flashed by the slot at different speeds by means 
of different tensions of a spring or a rubber band. Even 
gravity may be employed by acting on a falling shutter 
or on a pendulum raised to different vertical heights. 
For such rapid movements it is better to have the E sta- 
tionary and to expose it by the passage of a slot in an opaque 
but movable cover. This is shown in Fig. 46. An " en- 
velope " of stiff cardboard is made with a slot in the front 
as a, and a letter E on the back c, but in a position such that 
it shows in the slot when the slide b is removed. The slide 
b has a slot which passes the slot in the cover when the 
slide is as far in the slot as it will go. This is assembled 
by placing b in the slot and putting two rubber bands R 
around the assembly as shown. The exposure of the letter SPEED OF VISION 
197 
E is less the further out the slide b is pulled against the 
rubber bands. 
An improvement in some respects over the method em- 
ploying a single letter is achieved by having a long slot 
several times wider at one end than the other as shown in 
d, Fig. 46. Several similar letters are exposed simultaneously 
b 
E 
[- 
R 
R 
a 
Front 
b 
Slide 
c 
Back 
ab-c-Assembly 
eXa1e e e 
d 
Fig. 46. A simple device for demonstrating the effect of intensity of illu- 
mination on the time required to distinguish an object, a is a slot in the 
front of an envelope and a letter E is on the back, c. The slide b is inserted 
and two rubber bands R are slipped over as shown in the assembly. By 
pulling the slide out against the bands different exposures of the E can be made. 
d shows a wedge-shaped slot in which several E's can be exposed for different 
periods simultaneously. 
but for different periods of time. Thus a demonstration of 
the time element can be made without varying the intensity 
of illumination. At the proper intensity of illumination 
the letter at the widest part of the slot (longest exposure) 
will be clearly distinguished while the letter exposed the 
shortest time will not be distinguishable. 
Mr. G. H. Stickney devised a simple demonstration by 
standing a cardboard cylinder on the turn-table of a modern 198 
LIGHT AND WORK 
phonograph. Letters of various size are pasted or printed 
on this cylinder. The speed of the turn-table can be varied 
and also the intensity of illumination. In the latter case 
the low intensity can be that contributed by the lighting 
of a room at a desirable point. A reflector containing a 
lighted lamp can be held in the hand and on suddenly 
directing the light on the cylinder, some of the letters, which 
could not be recognized at the low intensity, suddenly 
become visible. 
Obviously the shutter of a camera can be employed for 
the purpose of demonstrating speed of vision. The back 
of the camera may be removed and also the lenses and the 
the test-object may be viewed directly through the lens 
aperture. The period of exposure can be varied in the 
usual manner. Sometimes it may be better to view the 
image of the test-object on the ground glass of the camera 
with the lenses in their usual position. In fact, if the 
shutter is dependable a real investigation may be conducted 
in this manner but, of course, the great reduction in lumi- 
nosity of the image compared with the luminosity of the 
object must be taken into account.  
Representing the result of indifference to lighting. 
(Above) The average illumination intensity was 0.2 foot-candles. The in- 
adequate illumination from bare lamps and improper reflectors results in 
lowered production, decreased safety, visual discomfort, and general depression 
of workmen. 
(Below) The same factory after rehabilitating the lighting system. The 
average illumination intensity is now from 6 to 10 foot-candles. Chapter X 
LIGHTING AND PRODUCTION 
ANY factor of our work environment which is essential 
to efficient human activity, is an important influence 
upon the output of a worker. This is true in home, office, 
factory, and in every place where vision is involved. How 
plainly we can see our work and our surroundings deter- 
mines the certainty with which we act. How quickly we 
can distinguish things determines to a great extent the speed 
with which we perform the entire act. Seeing plainly and 
quickly makes for increased safety. A flood of light, 
properly controlled, is stimulative. It is not a powerful, 
forceful stimulant, but it is an enlivening factor. The sur- 
roundings are more cheerful and the worker's general 
attitude toward his work, his environment, and even toward 
life itself is bound to be influenced favorably. Certainly 
light and lighting are important in all these respects. The 
most effective lighting is that which makes it possible to see 
and to act with the greatest certainty, speed, safety and 
comfort. This will likely be attained when the important 
characteristics of daylight outdoors shall have been brought 
indoors. That is, light of the quality or spectral character 
of noon sunlight and illumination intensities of hundreds 
of foot-candles eventually may be found in specifications 
for the most effective lighting. There are many other fac- 
tors such as diffusion, distribution of light, and brightnesses 
of surroundings which must be considered, but, in general, 
we may follow nature in this respect, in the absence of 
further knowledge. 
It is interesting to note that increased production of 
workers due to better lighting is not obtained at the expense 
199 LIGHT AND WORK 
200 
of the worker's energy. Good lighting does not force a 
worker to do more work in a given time. By increasing the 
ease or certainty of seeing it makes work easier and acts more 
certain. It eliminates the waste of time which is inevitable 
under improper and inadequate lighting. The worker is 
working just as hard (and with greater annoyance) under 
poor lighting as under good lighting but he is not accom- 
plishing as much. 
What May Be Expected from Improved Lighting 
When we introduce the factor of cost we must modify, 
if necessary, our specification of the most effective lighting. 
There is no doubt that increased intensity of illumination 
and a general improvement in most lighting systems pay 
very large dividends. However, it is obvious that there is 
a point where the law of diminishing returns demands notice. 
Fortunately, great improvements in indoor lighting are still 
profitable rather than costly. That is, the value of the 
results of the improved lighting is generally far in excess 
of the increased cost of lighting. In many cases, great im- 
provement in lighting can be achieved by replacing an 
ineffective and inefficient system with a modern installation 
without a comparable increase in cost of lighting. This is 
due to the fact that many present systems are wasteful of 
light and are not productive of the best conditions for see- 
ing. The problem before the lighting expert is to determine 
exactly what proper lighting is and to apply it to the point 
of adequacy compatible with its cost and returns. In other 
words, the optimum conditions of lighting and the value 
of its results must be determined. Investigations so far 
indicate that we are far from the point where we must 
begin to consider the law of diminishing returns. All laws 
pertaining to vision, as shown in Chapters VIII and IX, 
show an increase in ability to see and to work with increase 
in illumination intensity, assuming other factors in lighting LIGHTING AND PRODUCTION 
201 
to be satisfactory. Tests of production under increasing 
intensities of illumination, indicate a tendency of produc- 
tion to continue to increase at still higher intensities than 
the maximum used in the investigations. 
Some of the advantages of better lighting can not be 
accurately appraised in terms of dollars, but all will agree 
that they are worth something. The lighting expert has 
known what the principal advantages of improved lighting 
should be. As a consequence, in a survey of artificial light- 
ing conditions in factories made a few years ago, manufac- 
turers were asked to rank the various advantages as they 
saw them. Doubtless they did not possess exact records of 
the results of improvements in lighting, but their attitude 
in general is of interest. The percentage of manufacturers 
interviewed who included the various principal advantages 
known to the lighting expert were as follows: 
Increase in production 
Decrease in spoilage 
Decrease in accidents 
Percent 
79 
7i 
60 
Improvement in discipline 
5i 
Improvement in hygienic conditions 
4i 
This discussion is confined chiefly to manufacturing but 
it should be noted that the principles apply also to offices, 
stores, traffic, and in all places where there is human activity. 
The productiveness of better lighting in offices and in stores 
can be measured in terms of dollars as well as in factories. 
In 1919, about 9,100,000 wage earners were working in 
manufacturing industries in the United States. Besides these 
there were about 1,447,000 salaried employes and about 
269,000 proprietors and firm members. The total persons 
engaged was 10,812,736. Many of these were office-work- 
ers. With only one-tenth of the population of this country 
working in manufacturing industries, it is obvious that 
many millions are engaged in other activities and that a 202 
LIGHT AND WORK 
large percentage of these are in offices. Here good lighting 
also yields satisfactory dividends as the reading tests de- 
scribed in Chapter IX and other observations prove. 
Effect of Improved Lighting on Production 
Although analyses of visual processes, and the relation of 
lighting to them, are exceedingly important in learning how 
to use light, a final factor of great interest is the increased 
production, if any, attending an improvement in lighting. 
The effects of better lighting in increasing visibility and 
speed of vision are now well established beyond question. 
Much more work needs to be done before a complete science 
of lighting will be available. However, in the meantime 
investigations in factories have proceeded in parallel with 
laboratory researches. The actual results of improved light- 
ing on production can be measured, although such an in- 
vestigation, to be convincing, must be conducted in a more 
or less elaborate series in which all influential factors are 
either controlled or their influences determined. 
Some interesting results were obtained by W. A. Durgin 
several years ago by determining the production of various 
departments in several factories before and after the re- 
habilitation of the lighting. His investigations were more 
of the nature of a reconnaissance, nevertheless, they were 
extremely valuable. A brief summary of a few tests made 
by W. A. Durgin on the influence of improved lighting on 
production in factories is presented herewith. 
Case I. An increase in intensity of illumination from 4 
foot-candles to 12 foot-candles resulted in increases in 
various departments as high as 27 per cent and as low as 8 
per cent. The average was 15 per cent. The original light- 
ing was much better than usually found in factories; still 
there was a marked increase in production when it was 
improved. 
Case II. The original lighting was very unsatisfactory, LIGHTING AND PRODUCTION 
the system (if it could be called such) consisting of bare 
lamps and dropcords. The intensity of illumination was 
increased 25 times with an increase in consumption of elec- 
trical energy of only 7 times. The result of the improved 
lighting was an increase in production in one department as 
much as 100 per cent. In no department was the increase 
in production less than 30 per cent. This was an extreme 
case. 
Case III. The intensity of illumination was increased 
from 3 foot-candles to 12 foot-candles with an accompany- 
ing increase in consumption of electrical energy for lighting 
of only 3 times. The increase in cost of lighting was equiva- 
lent to 1.2 per cent of the payroll but the increase in 
production averaged 10 per cent. 
Case IV. The intensity of illumination was increased 6 
times but the consumption of electrical energy for lighting 
was only doubled. The increase in production averaged 
10 per cent. 
Durgin concluded that the average result of increasing 
the intensity of illumination to a reasonable value and of 
properly designing the lighting installation in factories, 
would be at least a 15 per cent increase in production at an 
actual cost of lighting equivalent to not more than a few 
per cent in the payroll. Many rehabilitations of artificial 
lighting have been achieved with great improvement of the 
lighting at a cost of about one per cent of the payroll of all 
the workers using the lighting. 
An interesting case had been reported by John Magee 
in a department of a factory turning out piston-rings. The 
investigation of the influence of lighting on production ex- 
tended over fifteen months. The original lighting system 
produced an average intensity of illumination of about 1.2 
foot-candles. Three levels of illumination from the re- 
habilitated lighting system were used in the investigation, 
averaging about four months each. The average results 
203 204 
LIGHT AND WORK 
during the tests at four different intensities of illumination 
are as follow: 
Foot-candles 
Per cent increase 
in production 
1.2 
0 
6-5 
13-0 
9.0 
17-9 
14.0 
25-8 
The maximum increase in cost of lighting was about 48 
per cent of the cost of the original lighting. This was 
equivalent to 2 per cent of the payroll. Savings represented 
by less spoilage and fewer accidents went a long way toward 
paying for the lighting. 
There are so many influences on production that, to be 
entirely convincing, an investigation of the influence of 
lighting on production must be extended over a long period 
during which careful control and observation of various 
factors must be exercised. An investigation recently con- 
ducted by D. P. Hess and Ward Harrison 43 is particularly 
interesting because of the elaborateness of the tests which 
extended over a period of ten weeks. An inspection de- 
partment of a factory was chosen and during the period 
the total number of inspectors varied from 38 to 48. The 
nature of the work consisted of inspecting material as it 
was turned out by automatic screw-machines and before 
heat-treating. The material consisted of various sizes of 
cups, cones, and threaded cones which are separate parts of 
a roller bearing. During the test-period over 7,000,000 
pieces of material were inspected. The work is carried on 
in three stages. The first group of inspectors gauge the 
material for diameter and depth; the second inspect for 
defects such as chatter, tool marks, ingot breaks, thin ribs 
and bad chamfer; and the third group for imperfections 
in the thread on the inside of the cones, bad mill, bad cham- 
fer and inadequate burnishing. Some of the work requires LIGHTING AND PRODUCTION 
205 
close visual inspection and in some of the gauging relatively 
little visual scrutiny is required. The inspectors were paid 
on an hourly basis. Relative humidity and temperature 
were maintained practically constant. 
The original system of artificial lighting contributed an 
average illumination of 2 foot-candles, but the distribution 
of the light was uneven and caused bad shadows. Daylight 
entered through distant windows and overhead skylights. 
During the first two weeks of the tests the original arti- 
ficial lighting and daylighting were used in combination, the 
500 
Old lighting 
Near lighting sqstem 
PIECES INSPECTED PER PERSON PER HOUR 
400 
Average pieces inspected per person per hour 
300 
FOOT-CANDLES 
200 
20 
15 
10 
5 
0 
100 
Average foot-candles on inspection table 
0 
TEN CONSECUTIVE WEEKS 
Fig. 47. Showing the average number of pieces of material inspected per 
person per hour for each of ten consecutive weeks; also the corresponding in- 
tensities of illumination for the various weeks. During the first two weeks 
the old lighting system (half daylight) was in operation. During the remain- 
ing eight weeks the new system of artificial lighting was operated. 
average intensity of illumination being 5 foot-candles. 
Then the new lighting system went into operation and the 
skylights overhead were blackened. The new system con- 
sisted of glass-steel diffusers located on 8-foot by 10-foot 
centers and were hung 12 feet above the floor. The illu- 
mination intensity from the new system was maintained 
constant for the week; then it was changed to another value. 
The actual illumination intensities employed were 5 foot- 206 
LIGHT AND WORK 
candles for the first two weeks from the original lighting 
system, and 6, 13, and 20 foot-candles from the new system. 
The intensities from the latter system differed each week 
and in no progressive order. The production was reduced 
for each week to the average number of pieces inspected 
per person per hour. In Fig. 47 this average production 
is shown for each week along with the intensity of illumi- 
nation. One will note at once that when the intensity was 
increased the production increased, and vice versa. This is 
very striking and convincing. 
lighting *~New lighting -> Neuf "9ht,n9 ~* 
5 foot-candles (half doglighij 
SJ|o 
(5 foot-candles (half doglight) 
6 foot-candlw 
13 foot-candles 
ZO foot candlei 
4 °/» 
(20 foot-candles) 
(6 foot- candlw)IJ 
(13 foct-condles) 
Z.4*/» 
o.yl' 
0.8°/. 
1.6 °/o 
PER CENT INCREASE IN PRODUCTION COST OELIGHTING IN PER CENT Of PAYROLL 
Fig. 48. A graphic illustration of the per cent increase in production 
and of the cost of lighting in terms of the payroll for the various intensi- 
ties of illumination. The production increases are averaged from Fig. 47. 
The average increase in production under various light- 
ings as well as the cost of light in per cent of the payroll, 
are presented in Table XII and Fig. 48. These data show 
not only that production increases as the intensity of illu- 
mination increases (other factors being practically constant) 
but also the advantage of a well-designed lighting system 
over a poorly designed one. This is shown by the consider- 
able increase in production for the new system with 6 foot- 
candles over that of the old system at 5 foot-candles. 
Reducing the former production to what it would very likely LIGHTING AND PRODUCTION 
207 
be at 5 foot-candles we find that about 3.4 per cent increase 
in production arises from the change-over from a system of 
poor design to one of good design, illumination intensity 
remaining constant. 
TABLE XII 
Showing the Relation of Intensity of Illumination to Factory Production. 
New Lighting System 
Foot-candles 
6 
13 
20 
Average production (pieces inspected per person 
per hour) * 
424 
440 
458 
Per cent increase in production 
4.0 
8.0 
12.5 
Cost of lighting in per cent of payroll 
0.8 
1.6 
2.4 
Per cent increase in production costing nothing.. 
3-2 
6.4 
10.1 
Ratio of increase in production to cost of lighting 
(per cent of payroll) 
5 
5 
5-2 
* Under the original system of poor artificial lighting and natural lighting giving 
an average illumination of 5 foot-candles the first two weeks test gave a production 
of 408 pieces inspected per person per hour. The cost of this poor artificial lighting 
(exclusive of the cost of natural light) was 0.3 per cent of the payroll. 
In order to visualize the gradual increase in production 
for a modern system of lighting as the intensity of illumi- 
nation is gradually increased, the data in Table XII and Fig. 
48 are plotted in Fig. 49. It is interesting to note how the 
three sets of data (corresponding to the three illumination 
intensities respectively 6, 13, and 20 foot-candles) lie on 
straight lines. These straight lines if projected backward 
would pass close to zero although of course production would 
cease when there would still be some light but not enough 
to do the work. The more interesting speculation is that 
connected with projecting the straight-line relations forward. 
(See Figs. 62, 63, 65.) There is no indication at 20 foot- 
candles of a diminution of the rate of increase in production 
with increase in foot-candles. This would begin to be notice- 
able when the maximum speed of muscular activity or of 208 
LIGHT AND WORK 
movement was approached or when the maximum speed of 
vision (discrimination or recognition) was approached. It 
is difficult to state which would be the first to begin to retard 
the rate of increase in production with increasing illumina- 
tion intensity. In fact, this depends upon the character 
of the work, for in different kinds of work these factors 
are differently involved. In work demanding rather fine 
14 
IZ 
10 
8 
6 
A 
2 
0 
%- 
V5 
X 
V 
PER CENT 
Value of increased production 
which actually costs nothing 
5 
FOOT- CANDLES 
10 
15 
ZO 
Fig. 49. Showing the relation of illumination intensity to the increase 
in production and to the cost of light in terms of the payroll. 
visual discrimination, the maximum speed of vision is not 
approached until high intensities of illumination are reached 
and even if the maximum speed of movement has already 
been obtained, production will continue to increase until 
there is no increase in the ease or speed of visual discrimi- 
nation. 
In Chapter XI the results of cleaning glass areas of day- 
lighting equipment in two cases, indicate that production 
was increased 5 per cent and 10 per cent respectively. LIGHTING AND PRODUCTION 
209 
Laboratory tests of the speed of reading black print on 
white paper showed that the average speed of a large group 
of observers increased 15 per cent when the illumination 
intensity changed from 4 foot-candles to 16 foot-candles. 
The speed of reading black print on gray paper (23 per 
cent reflection-factor) increased 50 per cent over the same 
range of illumination. Although this test was performed 
in the laboratory, it is a real production-test for it applies 
directly to a large number of workers in offices, etc. and 
the results strengthen our conclusions in regard to the effect 
of illumination intensity on factory production. Strictly 
laboratory researches of visual laws discussed in Chapters 
VIII and IX, also indicate roughly what we might expect 
in any activity involving vision. From all these data it 
appears very conservative to conclude that the output of 
our factories could be increased a net 10 per cent by better 
lighting without cost. Where the lighting is particularly 
faulty this increase would be much greater. It should be 
understood that, up to very high intensities, ease and, there- 
fore, speed of discrimination of finer details, at least con- 
tinues to increase as the intensity of illumination increases. 
This increase at the very high intensities of illumination 
is more for low-contrasts than for high ones. This is 
discussed further in Chapters XIII and XIV. 
It should be borne in mind that throughout this discussion 
the benefits of only increased production have been dis- 
cussed. Decreased spoilage, increased safety, better 
discipline, and less eye-strain and consequent discomfort 
are all accompaniments of better lighting. All these have 
monetary value but it is difficult to appraise them. We shall 
let them represent good measure after we have appraised 
the value of increased production. 210 
LIGHT AND WORK 
A Glimpse of Manufacturing Statistics 
If we accept the proof that improved lighting can create 
value out of proportion to the cost of the lighting; that is, 
that it can create net value, it is of interest to show what the 
increased production means for all the manufacturers in the 
United States. Certainly the lighting of nearly every factory 
can be improved somewhat either in intensity or distribu- 
tion of illumination. As pointed out in Chapter VII, only 
9 per cent of the factories visited in an extensive survey 
had what was termed excellent lighting as judged by the 
low standard a few years ago. Standards have risen con- 
siderably during these few years and recent data indicate 
the desirability of much higher intensities, so that it appears 
safe to say that even those factories could have the lighting 
improved and paid for by the increased production. 
In order to obtain a picture of our industrial economics 
some statistics have been condensed and rearranged from the 
Census of Manufactures for 1919, the last census available. 
In Table XIII a summary is given for the United States for 
the years 1904, 1911, 1914, 1919. It is seen that lighting 
in the industries affected over 9 million workers who turned 
out finished products valued at 62 billion dollars in 1919. 
In order to make it easier to compare various items, they are 
also given in per cent of the 1904 value in each case. The 
actual values are expressed in millions of dollars. 
Most manufacturing is found in cities where, owing to 
smoke and congestion, natural lighting is generally less 
than in smaller places. On the other hand, electric service 
is likely to be better and higher standards of artificial light- 
ing are more readily sold to factory management in the 
larger cities. Some statistics of manufactures for the year 
I9I9 pertaining to cities of 10,000 population and over and 
also to cities of 100,000 and over are presented in Table 
XIV. These figures are percentages of the total for the 
United States for the year 1919. LIGHTING AND PRODUCTION 
211 
Summary of Statistics of Manufactures for the United States. 
Actual values expressed in Millions of Dollars; 
Relative values in per cent of 1904 value. 
TABLE XIII 
1904 
1909 
1914 
1919 
No. of establishments 
216,180 
268,491 
275,791 
290,105 
Per cent 
100 
124 
127 
i34 
No. of wage earners 
5,468,383 
6,615,048 
7,036,247 
9,096,372 
Per cent 
100 
121 
129 
166 
Capital: 
Millions of dollars 
12,676 
18,428 
22,791 
44,467 
Per cent 
100 
i45 
180 
35° 
Wages: 
Millions of dollars 
2,610 
3,427 
4,078 
io,533 
Per cent 
100 
131 
156 
400 
Cost of materials: 
Millions of dollars 
8,500 
12,143 
14,368 
37,376 
Per cent 
100 
i43 
169 
440 
Value of finished products: 
Millions of dollars 
14,794 
20,672 
24,246 
62,418 
Per cent 
100 
140 
165 
423 
Value added by manufacture: 
Millions of dollars 
6,294 
8,529 
9,878 
25,042 
Per cent 
100 
136 
i57 
405 
Statistics of Manufactures for cities of 10,000 and over and for cities of 100,000 
and over in percentage of the total for the United States for the year 1919. 
TABLE XIV 
Cities 
10,000 Population 
and Greater 
Cities 
100,000 Population 
and Greater 
Number of factories 
57.2 per cent 
38.7 per cent 
Number of wage-earners 
70.2 
4i.7 
Population 
42.3 
25-9 
Value of finished products 
72.6 
45-7 
Value added by manufacture 
74.1 
46.3 
From Table XIV it is seen that about three-fourths of 
the factory wage-earners and the manufactures are in cities 
of 10,000 inhabitants and greater, and at least one-half of 212 
LIGHT AND WORK 
the factories (measured by wage-earners and output) are 
located in large cities - in other words, within the reach of 
the larger and better organized central-stations, electrical 
contractors, etc. 
Having had a glimpse of statistics of manufactures in 
Table XIII, it is interesting to see what even small increases 
in production amount to in the manufacturing industries 
alone. Taking the statistics of the Census Bureau for 1919, 
one per cent increase in production due to improvement in 
lighting above the additional cost of lighting is, for the year, 
equivalent approximately to 
100,000 wage-earners 
$100,000,000 wages 
$250,000,000 value added by manufacture 
$625,000,000 value of finished products 
It seems to be fairly well established that reasonable 
increase in intensities of illumination and rehabilitation of 
lighting systems would on the average result in a net in- 
crease in production of at least io per cent. On this basis 
the improvement in factory lighting at the present time, 
would yield a stupendous result each year. It would do 
the work of one million wage-earners at a saving of one 
billion dollars in wages. This equivalent of a million wage- 
earners would increase the total value of finished products 
by 6 billion dollars. Without any additional cost, there 
would be an increase each year of over 2.5 billion dollars in 
the value of finished products added by manufacture at 
the present rate of manufacture. In other words, proper 
attention to lighting throughout the United States would 
create, on a basis of production in 1919, a net value of over 
2.5 billion dollars exclusive of decreased spoilage, labor 
turnover, and accidents. 
Thus improvement in lighting in 1919 could have been 
equivalent to 3 times the wage-earners employed in the LIGHTING AND PRODUCTION 
manufacture of automobiles that year, and it could have 
increased (free of cost) the value added by manufacture to 
the value of finished products, an amount equivalent to the 
entire value of automobiles manufactured in 1919. In 
other words, if we had proper and adequate lighting through- 
out our factories, we would get the equivalent of our auto- 
mobiles without cost. By doing the same for our offices 
and other work-places, we would likely get the equivalent of 
the necessary gasoline and oil to operate them. 
213 
o 
•b 
G> 
OU 
3 
<F 
BILLIONS OF DOLLARS SAVED OR VALUE CREATED 
4.0 
3.5 
3.0 
Z.5 
2.0 
1.5 
1.0 
0.5 
PER CENT INCREASE IN PRODUCTION 
rM 
-- 
I 
Fig. 50. Showing the billions of dollars of value which could be 
created without cost each year and also saving in labor cost for various 
increases in production due to improved lighting throughout the manu- 
facturing industries of the United States. These values are based on the 
1919 census of manufactures. 
In computing the value in manufactured products which 
could be created by improvement in lighting, only the value 
added to finished products by manufacture has been con- 
sidered. That is, the value of the materials has not been  LIGHT AND WORK 
214 
taken into account. This is not affected excepting by de- 
crease of spoilage and we have no accurate figures in regard 
to the effect of improved lighting on that item. Spoilage 
is generally considered to be decreased about the same 
percentage that production is increased. In order to show 
the value which could be created by adequate and proper 
lighting each year at the 1919 rate of manufacture, Fig. 
50 is presented. The created value is given in billions of 
dollars for various increases in production up to 16 per cent. 
If this increase in production is not desired, the improve- 
ment in lighting can be appraised in terms of saving in labor 
cost. This is also given for the same range of increase in 
production effected. The difference between these two items 
is the value added by increased production of that portion 
of the manufacturing process which does not include wages, 
such as equipment. However, it does not necessarily follow 
that all other links in manufacturing increase their output 
in the same proportion as the workman does; but the human 
element is so interwoven with manufacturing that perhaps 
we are not far wrong in assuming that it does. If we are 
wrong at least the saving in labor cost still holds. Chapter XI 
THE VALUE OF PROPER MAINTENANCE OF 
LIGHTING SYSTEMS 
STRIKING evidence of a lack of appreciation of the 
value of adequate light is seen by the prevalence of 
dirty glass areas in factories. The daylight equipment in 
most factories represents an appreciable investment from 
which the maximum return is not obtained unless the glass 
is kept reasonably clean. Manufacturers who would not 
think of neglecting other essentials in profitable production 
will let the natural lighting diminish month after month 
by the accumulation of dirt. The loss from such neglect is 
not plainly visible but it is a loss none the less. With 
adequate artificial lighting the loss due to diminished day- 
light is greatly reduced because the cost of artificial light 
is a very small part of the total cost of production. How- 
ever, manufacturers that are so unappreciative of the value 
of adequate light are not likely to be among those relatively 
few who have installed high-standard artificial lighting. 
We wash the windows in our homes and offices through 
pride, if for no other reason, before the decrease in light 
becomes serious. In our factories this incentive is too often 
absent. As a result, " day " indoors is materially shortened 
but still worse is the loss of production, the decrease in 
cheerfulness, the lowering in discipline and in orderliness, 
and doubtless indirect losses such as increased labor turn- 
over. The same results obtain from neglect of artificial 
lighting equipment. The accumulation of dirt on lamps, 
reflectors, ceiling and wall is costly. Improper lamps, 
burned out ones, voltage below normal, and old lamps de- 
crease the output of light. 
215 216 
LIGHT AND WORK 
Effect of Cleaning Glass of Daylight Equipment 
After the glass areas in a factory have been cleaned the 
management is usually convinced of many advantages such 
as increased production, less spoilage, a decrease in number 
of accidents, better discipline, fewer mistakes, and more 
cheerful condition. Usually the records are not adequate 
to reduce these items to monetary value and, indeed, some 
of them cannot be. However, it is generally conceded 
through experience and observation, that the benefits of 
cleaning glass areas greatly exceed the cost. Data supplied 
by the management of two different factories are presented 
herewith. 
Case I. In the galvanizing department of a large manu- 
facturing company the natural lighting had diminished 
markedly by the deposit on the windows. This deposit re- 
sisted various attempts at cleaning them in which sand, lye, 
acids, etc. were used. However, a cleaning compound was 
finally found which did the work. Brief statistics of the 
cost and of the value of cleaning are given herewith. 
Cleaning compound $30.00 
Labor (11 days) $66.00 
Total for 6,000 sq. ft $96.00 
Per square foot $ 0.016 
Costs oj Cleaning 
Value oj Cleaning 
Saving in artificial lighting per day $ 0.60 
Increase (5%) in labor efficiency $T3-75 
Saving per day $14.35 
Saving per month $358.75 
Increase in profit on extra production (not given) 
Saving due to better inspection (not given) 
The saving in labor cost represented in an increase of 
5% in labor efficiency in one week paid for the cleaning PROPER MAINTENANCE OF LIGHTING SYSTEMS 
217 
without taking into consideration the profit on the extra 
production, the value of better inspection (in this case errors 
of inspection were very costly) and other factors. The 
management stated that" one of the main reasons for clean- 
ing the skylights was that it added greatly to the general 
appearance of the plant. This alone is worth the expense 
required." 
Case II. In a warehouse of a very large manufacturing 
organization, 16,000 square feet of skylight were cleaned. 
Keeping this glass clean is a difficult matter in this plant 
because of the proximity to chemical industries; however, 
the management experimented until a cleaning compound 
was found that would do the work well and at a reasonable 
cost. Brief statistics of the cost and value of cleaning the 
glass are given herewith. 
Costs oj Cleaning 
Cleaning compound $250.00 
Labor (448 hours) $246.40 
Total or i63ooo sq. ft $496.40 
Per square foot $ 0.031 
Saving in artificial lighting $ 150.00 
Increase (5%) in production $7500.00 
Gross annual saving $7650.00 
Net annual value $7153.60 
Reduction in errors 5% 
Reduction in accidents 10% 
Value of Cleaning per year 
The management of this warehouse state that " with 
better light the spirit and atmosphere of the plant changed. 
The workmen found the conditions far more pleasant and 
became even better satisfied with their jobs. Eye-strain 
was greatly reduced. On the whole the effect on the work- 
men was most gratifying." The cleaning of the glass areas  LIGHT AND WORK 
218 
paid for itself 28 times on the basis of cleaning once a year 
without considering the monetary value of a decrease of 5 
per cent in errors in filling orders and a decrease of 10 per 
cent in accidents. Although no attempt was made to 
appraise the effect on labor turnover in this department 
employing 100 men, the management believes that " the 
more pleasant working conditions affect the labor turnover 
favorably." 
There is no doubt that, where conditions are favorable, 
daylight will be continued to be used. It is an ideal illumi- 
nant, but owing to the difficulties of getting it indoors in 
sufficient quantity and of proper distribution, the design of 
daylighting equipment demands skill. That daylight in- 
doors costs something is not at all to its discredit; however, 
this must be recognized if lighting is to receive just con- 
sideration. After daylighting equipment is installed at 
considerable cost it should be maintained or it does not pay 
dividends. 
Absorption of Light by Dirty Glass of Daylight 
Equipment 
In Chapter III it was shown that clean crystal glass 
transmits from 75 to 90 per cent of the incident light. In 
factory districts, or for that matter in any part of our cities, 
glass accumulates a film of dirt rapidly. Particularly in 
factory districts this is very difficult to clean if neglected 
for a long time. However, there are cleaning compounds 
to which any carbon scale or any dirt will yield. In Table 
XV are shown the transmission-factors of seven different 
glasses just as they were taken from skylights and win- 
dows and also after they were cleaned by means of a 
special cleaning compound. By cleaning, their transmis- 
sion-factors were increased enormously, as seen by the last 
column. In one case the glass was so dirty as to be prac- 
tically opaque. In the other cases the lighting had been PROPER MAINTENANCE OF LIGHTING SYSTEMS 
219 
reduced by the accumulation of dirt to an average of about 
one-sixth of the original value with clean glass. In terms 
of loss in production this means many millions of dollars 
in this country per year. 
Results of Cleaning Glasses in Factory Daylighting Equipment. 
TABLE XV 
Type of glass 
Transmission factor (per cent) 
Transmission 
Increased by 
Cleaning 
Before cleaning 
After cleaning 
Plain clear 
12 
88 
7.3 times 
Plain clear 
16 
88 
5-5 
Fine ribbed 
14 
80 
5-7 
Fine ribbed 
17 
76 
4-5 
Fine ribbed 
°-3 
75 
250-0 
Wavy wire glass 
13 
80 
6.1 
Wavy wire glass 
5 
75 
150 
Measurements of intensities of illumination before and 
after cleaning the glass areas yield startling results. In 
many cases the increase is as much as 5 to io times and it 
is not uncommon to increase the illumination intensity in 
some parts of the interior from an average of one foot- 
candle with dirty glass to as high as 15 foot-candles. 
Depreciation of Artificial Lighting 
All equipment is exposed to dust and it is to be expected 
that artificial lighting equipment does not escape. The rate 
of depreciation in output of light due to accumulation of 
dust depends very much upon the location and type of light- 
ing equipment. A direct-lighting system consisting of dense 
diffusing glass shades, opaque reflectors, or non-ventilated 
diffusing glass units will not depreciate in light output as 
rapidly as most other systems under similar conditions. It 
used to be thought necessary to ventilate lighting units. 
However, with modern tungsten lamps no ventilation is 220 
LIGHT AND WORK 
necessary if the enclosing glass is as large as it should be to 
provide a sufficiently low Ventilating orifices 
suck in the dust along with the air and much of it remains 
deposited on the lamp and various parts of the equipment. 
In from one to three months, depending upon the location 
and lighting equipment, a depreciation in light output from 
io to 50 per cent can readily take place from the accumula- 
tion of dirt alone. A regular schedule of maintenance is 
found to be the best method of insuring against great loss 
Final Intensity 
1 
6 
5 
3 
2 
1 
0 
Increase:- 
Walls and Ceiling 
Refinished 
Final Intensity 
7 
6 
5 
3 
2 
I 
0 
Increase:- b° 
Walls Refinished 
Increase:- 
New Lamps of g 
Proper Voltage S 
FOOT-CANDLES 
Final Intensity 
Increase:- 
Reflectors Cleaned m 
Increase:- , 
Lamps and Re- h* 
Hectors Washed 
Increaser- 
Lighting Units X 
Cleaned 
Intensity 
as found 
Intensity 
as found 
Increase:- • 
Burnouts in 
light Fixtures 1 
Renewed 
Intensity 
as found 
Occasional Maintenance 
CASE 1 
CASE JI 
No Maintenance 
CASE 111 
Periodic but Inade- 
quate Maintenance 
Fig. 51. Chart showing the importance of prompt renewal of burned-out 
lamps and systematic cleaning of the lighting equipment. 
of light. Although the depreciation of artificial lighting 
equipment is not generally permitted to progress to the ex- 
treme that is common with daylight equipment, it becomes 
very marked in many cases. This is particularly costly 
owing to loss of production and to other factors when light- 
ing intensities are initially quite low as is the case generally. 
The various factors and their influences upon light output 
and illumination intensities are well shown in Fig. 51 for 
three representative cases of semi-indirect and indirect 
lighting as determined by Ward Harrison. PROPER MAINTENANCE OF LIGHTING SYSTEMS 
221 
Case I. This is one of occasional maintenance. The 
intensity of illumination after a long period of neglect was 
2.7 foot-candles. The increase due to maintenance factors 
was as follows: 
Foot-candles 
Per cent 
increase 
Original intensity of illumination. 
.... 2.7 
- 
Washing lamps and reflectors.... 
.... 3.7 
37 
New lamps of proper voltage... . 
•... 5-i 
38 
Refinishing ceiling and walls 
.... 7.1 
40 
Total increase in illumination. .. . 
.... 4.4 
163 
Case II. This was one in which lighting equipment was 
not maintained. The results of maintenance were as 
follows: 
Foot-candles 
Per cent 
increase 
Original intensity of illumination. 
1.2 
- 
Burned out lamps replaced 
2.2 
80 
Lighting equipment cleaned 
.. .. 4.2 
90 
Total increase in illumination.... 
.... 3.0 
260 
Case III. This was one in which periodic but inadequate 
maintenance was carried out. The results of maintenance 
were as follows: 
Foot-candles 
Per cent 
increase 
Original intensity of illumination. 
.... 4.0 
- 
Lighting equipment cleaned 
.... 4.9 
23 
Walls refinished 
. ... 6.3 
29 
Total increase in illumination 
.... 2.3 
58 
In factories, direct-lighting systems are almost always 
specified, so that it may appear unfair to present mainte- 
nance data for semi-indirect and indirect systems. Ceilings 
and walls have much less effect on the illumination at the 
work for direct-lighting systems than for the others. The 
reader may drop this item out entirely if he desires, although 222 
LIGHT AND WORK 
in certain modern systems of factory lighting in which 
glass is a part of the equipment, the ceiling has an appreci- 
able effect on the illumination at the work. In fact, a ceil- 
ing moderately illuminated adds greatly to visual comfort 
and increases the cheerfulness of the interior. Furthermore, 
the conditions in factories are generally such as to cause 
more rapid depreciation in the lighting than those in offices. 
Anderson and Ketch 39 found lighting equipment exposed 
to oily sooty dust seven weeks depreciated more (from dirt 
alone) than when exposed to dry, fine dust for more than 
twice this period. Under these conditions they found depre- 
ciations for lighting equipment of the kind commonly used 
in factories varying from 11 per cent to 30 per cent due to 
accumulation of dust. Considering that in most factories 
lighting equipment is not regularly cleaned, that the periods 
between cleanings are generally quite long, that the reflec- 
tion from ceilings has some influence, that the voltage is 
not always proper, and that burned-out, blackened, or im- 
proper lamps are allowed to remain in the sockets, the 
total average depreciation is likely to be much greater than 
the value found by Anderson and Ketch. 
Cost of Depreciation in Lighting 
In the three cases given in the foregoing, the increase 
in intensity of illumination due to proper maintenance was 
58 per cent and 260 per cent respectively for the extreme 
cases. It is difficult to believe that something as important 
as a lighting installation will be permitted to depreciate so 
much through neglect, but the evidence is found on every 
hand. The chief trouble is that light is too commonplace. 
It is accepted as a matter of course and it receives little 
or no attention until the intensity of illumination becomes 
noticeably low. However, all this time there has been a 
loss of production, spoilage, wastage of time and perhaps 
accidents and lowering of discipline, all increasing unawares. PROPER MAINTENANCE OF LIGHTING SYSTEMS 
223 
In Chapter X it was shown that doubling the intensity from 
6 foot-candles to 12 foot-candles increased production 4 
per cent. Halving an already inadequate intensity of illu- 
mination is likely to reduce production much more than 
4 per cent. 
It is startling to think that neglect of lighting equipment 
in the United States is costing us a billion dollars per year; 
however, on the basis of what we know of the influence of 
lighting on production this is just what is taking place if the 
depreciation is sufficient to cause a decrease in production 
of 4 per cent. Considering that we begin with inadequate 
lighting and that it does not need to depreciate 50 per cent 
to be responsible for such a decrease in production, this 
estimate of a billion dollar loss per year is not unreasonable. 
A large factor in manufacturing is daylighting and it is a 
fact of common observation that daylight equipment is 
generally permitted to depreciate very much. 
The cost of cleaning direct-lighting equipment commonly 
used in factories is considered by A. L. Powell40 to average 
between 3 cents and 5 cents each. Obviously, cleaning 
equipment too often would cost more than the saving due 
to the better lighting. It is best to design the installation 
to provide the desired intensity of illumination after rea- 
sonable depreciation; that is, to provide a factor of safety 
in the initial intensity. After a period of depreciation, 
when the gain due to cleaning offsets the cost of cleaning, 
is obviously the time to clean the lighting units. The length 
of this period will vary widely with conditions and will 
depend upon the factor of safety allowed initially. The 
cost of depreciation is by no means merely the cost of light 
lost. In fact, the decrease in production and the losses due 
to other factors which we have associated with better light- 
ing are much greater than the cost of the light lost. 
There is a depreciation in the output of light from a 
clean electric filament lamp (and mercury arcs as well) with 224 
LIGHT AND WORK 
hours of operation. This is not very serious during the 
normal life for which the lamp is designed but many lamps 
remain in service long after their normal life. Such lamps 
may cause considerable depreciation as seen in Fig. 51. A 
profitable schedule of cleaning can be determined without 
much difficulty in any given case. A foot-candle meter, 
which can now be purchased at a reasonable price, should 
be a part of the equipment of the larger factories and office 
buildings. In any case, the management will find that de- 
terminations of illumination in foot-candles are instructive 
and valuable. By means of data in other chapters it is 
easy to relate illumination intensities to production. By 
converting production into profits it is certain that the 
management of any factory, office, store, or any place where 
persons work for pay and are paid for profit, will see that 
lighting, adequate and proper, is the most satisfactory and 
profitable entry on the payroll. Chapter XII 
THE LIGHTING VALUE OF PAINT 
OFTEN when we are trying to see something in deep 
shadow we hold our hand, a white envelope or some- 
thing else so that it reflects light into the shadow. This is 
a minor use of the lighting value of surfaces of high reflec- 
tion-factor. If the amount of light lost through neglect 
in painting surfaces or maintaining them of high reflection- 
factor could be computed, there is no doubt that the result 
would be staggering. This does not apply only to the use 
of paint but also to any surface covering; however, paint 
is the more practicable covering from the viewpoint of 
this book. The measurement of reflection-factor is now a 
comparatively simple matter in a properly equipped labora- 
tory but accurate results can only be obtained under such 
conditions. Usually color-difference is involved and results 
obtained outside the laboratory are far from accurate. 
Furthermore, the reflection-factor of a surface depends 
upon the direction and distribution of the incident light in 
all cases excepting that of a perfectly diffusing surface. In 
general, walls and ceiling play a part in the distribution of 
light to the work-plane but the part they play not only 
depends upon their reflection-factors but also upon the 
character of the system of lighting. In factories and other 
places it has been found that machinery and other equipment 
when finished in a lighter color than heretofore, contribute 
an appreciable amount of light to the work-places. Further- 
more, funereal finishes of walls, woodwork, and equipment 
are giving way to those which make the surroundings more 
cheerful. 
22$ 226 
LIGHT AND WORK 
Influence of Ceiling and Walls 
In lighting a certain work-plane, such as a horizontal 
one 30 inches above the floor, the ratio of the amount of 
light reaching this plane to the total light emitted by the 
light-sources is termed the utilization-jactor. Some of the 
light emitted by the light-sources is absorbed by the light- 
ing equipment because a reflector does not reflect all the 
incident light and a diffusing glass does not transmit all 
the light. The losses in clean factory-lighting equipment 
of approved designs average about 30 per cent. The influ- 
ence of the ceiling and walls obviously depends not only 
on their characteristics but also upon the character of 
the lighting system and upon the size of the room. In 
other words, they can influence the lighting only in so far as 
they have opportunity and this is a matter of the percentage 
of the total light which is incident upon them. For example, 
let us take a spherical room without windows or other open- 
ings and assume the ceiling, walls and floor painted 
uniformly. Also let us suppose a light-source to be hung in 
the center of this room and that the light-source be such 
that its direct light produces an illumination of one foot- 
candle on the walls. The light emitted toward the boun- 
daries of the room is reflected many times and the illumina- 
tion intensity of one foot-candle is augmented by this re- 
flected light. This augmentation is zero if the walls, ceiling, 
and floor were perfectly black but becomes appreciable even 
for low reflection-factors. The results for a spherical room 
are easily computed, so they are given in Table XVI for 
the assumptions just made. In the table it is seen that when 
the surrounding walls are perfectly black, the illumination 
of the wall is just that of the direct light - one foot-candle. 
As the reflection-factor of the walls increases the illumina- 
tion on the walls increases. When the reflection-factor is of 
about the best commercial white paint the illumination is THE LIGHTING VALUE OF PAINT 
227 
doubled and as the reflection-factor increases from this 
point the illumination rapidly increases. It would be in- 
finite if the walls were perfectly white, because no light 
would be lost at all by absorption and it would continue 
to be reflected and re-reflected. 
TABLE XVI 
Showing the Influence of the Reflection-Factor of the Walls of a 
Spherical Room upon the Illumination of the Walls. An Unshaded 
Light-Source Hangs in the Center of the Sphere of such Intensity that 
It Illuminates the Sphere with Uniform Direct Light to an Intensity of 
One Foot-Candle. 
Material 
Reflection-factor 
Foot-candles 
of Wall 
on Wall 
Perfect black 
o per cent 
1.00 
Ordinary black paint 
4 
1.04 
Dark gray 
IO 
1.11 
Medium gray 
5° 
2.00 
Commercial white 
80 
5.00 
White powder 
9° 
10.00 
Best white powder 
98 
50.00 
Perfect white 
100 
Infinite 
Although the foregoing is only a theoretical consideration 
it is interesting in illustrating the possibilities of reflecting 
light. In our interiors there are many objects which absorb 
light and even white walls and white ceiling are not over 
80 per cent reflection-factor at best. It is impracticable 
to attempt to keep the floors highly reflecting and, further- 
more, white surroundings are by no means the most desir- 
able conditions from the viewpoint of eye-comfort and the 
best vision. Ceilings should be of high reflection-factor 
if the system of lighting is such as to emit much light 
directly to the ceiling. When considered from all angles it 
is questionable whether we should have walls directly in 
the field of view of greater reflection-factor than 50 per 
cent. 
In the case of direct lighting with opaque white enamelled 228 
LIGHT AND WORK 
or silvered glass reflectors little light reaches the ceiling, so 
its reflection-factor from the viewpoint of artificial lighting 
is of little importance although it is of great importance in 
daylighting from windows. In the case of indirect lighting 
and so-called semi-indirect lighting a large percentage of the 
light is emitted upward to the ceiling and therefore the re- 
flection-factor is important. Walls are important to the ex- 
tent of the percentage of light that impinges upon them. 
Obviously in large rooms they are of much less importance 
than in small rooms. They are generally of much greater 
importance in natural lighting from windows than in arti- 
ficial lighting. 
In order to show the influence of the reflection-factors 
of ceiling and walls in artificial light, Table XVII is pre- 
sented for three different types of lighting systems and for 
three sizes of rooms. The lighting equipment is assumed 
to be hung io feet above the horizontal work-plane upon 
which the relative values of illumination from a certain 
number of lamps are shown in the Table. These data 
were deduced from the excellent work of E. A. Anderson41 
on utilization factors. The numbers in parentheses follow- 
ing the words " ceiling " and " walls " are the reflection-fac- 
tors respectively. The direct-lighting system is one in which 
approved white enamelled or silvered glass reflectors are 
used. The second system employs approved enclosed dif- 
fusing glass equipment. The indirect lighting system is one 
in which approved silvered glass reflectors or dense opal 
glass bowls are used. 
The data in Table XVII will bear considerable study. 
In the large rooms the walls play an inappreciable part and 
it is interesting to note in item a, for example, the effect of 
the walls as the room size diminishes. This is conspicuous 
in all the items. The ceiling plays a small part in the case 
of direct lighting with opaque reflectors. This is seen by 
comparing item b with item d where the walls have the same THE LIGHTING VALUE OF PAINT 
229 
TABLE XVII 
Showing Relative Foot-Candles on Work-Plane from a Given Number of 
Lamps Used in Three Different Lighting Systems for Various Reflection-Factors 
of Walls and Ceiling and for Three Sizes of Rooms. 
Lighting System and Various 
Reflection-factors of Ceiling 
and Walls 
Relative Foot-candles on Work-Plane 
Small room 
(10 ft. by 30 ft.) 
Intermediate 
room 
(20 ft. by 50 ft.) 
Large room 
(100 ft by 100 ft) 
Direct Lighting System: 
a. Ceiling (70) walls (50)... 
50 
90 
100 
b. Ceiling (70) walls (30)... 
42 
86 
97 
c. Ceiling (70) walls (10)... 
35 
81 
94 
d. Ceiling (30) walls (30)... 
40 
84 
94 
e. Ceiling (30) walls (10)... 
35 
81 
9i 
Enclosed Diffusing Glass: 
f. Ceiling (70) walls (50)... 
32 
67 
80 
g. Ceiling (70) walls (30)... 
25 
59 
74 
h. Ceiling (70) walls (10)... 
20 
54 
68 
i. Ceiling (30) walls (30)... 
20 
49 
58 
j. Ceiling (30) walls (10)... 
17 
45 
55 
Indirect Lighting System: 
k. Ceiling (70) walls (50)... 
23 
52 
65 
I. Ceiling (70) walls (30)... 
19 
46 
60 
m. Ceiling (70) walls (10)... 
16 
42 
55 
n. Ceiling (30) walls (30)... 
10 
25 
32 
0. Ceiling (30) walls (10)... 
9 
23 
29 
reflection-factors but those of the ceiling are 70 and 30 per 
cent respectively. 
Enclosed diffusing glass equipment emits a great deal of 
light upward to the ceiling and outward to the walls. This 
is quite evident by the generally lower values of relative 
illumination under this system than under the direct-light- 
ing system. Both the ceiling and the walls influence the 
relative foot-candle values on the work-plane. By compar- 
ing A g and the influence of the reflection-factor of the 
walls is seen. By comparing g and i or h and j the influence 
of the reflection-factor of the ceiling is seen. There are 
several outstanding features of Table XVII: (1) the appre- 230 
LIGHT AND WORK 
ciable influence of walls of small rooms or narrow rooms in 
which direct-lighting opaque reflectors are used: (2) the 
great influence of both walls and ceiling in all the rooms 
but particularly those of small and intermediate sizes in 
the case of systems of enclosing diffusing glass; (3) the 
very great influence of walls and particularly ceilings in the 
case of indirect-lighting systems. Surely Table XVII 
presents adequate proof of the lighting value of paint. 
Reflecting Characteristics of Surfaces 
Paints exhibit the various reflecting characteristics that 
surfaces do in general. These are shown in Fig. 52 where 
Ft 
1 
.A-'-'p 
Ft 
1 
a 
b 
c 
Fig. 52. Common reflection characteristics of surfaces: a, a regular or 
specular reflection: b, perfectly diffuse reflection: c, spread reflection. 
I is the incident beam of light and R is the reflected light. 
Regular or specular reflection a is that of a perfect mirror 
where the angle of reflection R equals the angle of incidence 
I. This is approached by fairly polished metals and glossy 
varnished surfaces. A thick coating of glossy varnish on a 
jet black surface is a very good mirror. Such surfaces are 
undesirable from the viewpoint of lighting, because they re- 
flect images of the light-sources, thereby causing glare and 
the attendant effects such as visual discomfort, eye-strain 
and reduced visibility. Glossy desk-tops, glossy walls and 
ceiling, and various polished surfaces are possible sources 
of glare. 
Paints with dull finish approach perfect diffuse reflection 
b fairly closely although semi-matte finishes reflect not only 
diffusely but superposed on this is a sort of regular reflection THE LIGHTING VALUE OF PAINT 
231 
which may be said to tend toward spread reflection c. A 
painted surface coated with glossy varnish or an enamelled 
surface may be represented by a combination of a and b; 
that is, by regular reflection superposed upon diffuse re- 
flection. A good example of this is a thin sheet of glass 
laid upon a white blotting paper. 
Metallic paints, such as aluminum paint, are excellent 
examples of spread reflection shown in c, Fig. 52. 
1.0 
0.9 
4 
0.8 
0.7 
0.6 
0.5 
0.4 
0.3 
0.2 
0.1$ 
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 "Vag 1.0 
'Per fed- 
Mirror 
Specular Reflecfion-Facfor 
do/ished S urfaces 
a 
Made Surfaces 
/Perfect 
(Black) 
(Gray) 
(Whi+e) 
Diffuse Ref lection-fa of or 
Fig. 53. Showing the amount of regular and of diffuse reflection of va- 
rious surfaces which exhibit a combination of these two types of reflection. 
The regular reflection is determined for perpendicularly incident light. 
A Glossy black paint 
B Matte black paint 
C Glossy gray paint 
D White blotting paper 
E Best matte white paint 
F Best glossy white paint 
G Glazed white porcelain 
H Powdered zinc oxide 
I A surface of glass or water 
J Two surfaces of thin glass 
K Polished silver 232 
LIGHT AND WORK 
So many surfaces show the characteristics of regular and 
diffuse reflection that it is interesting to plot them on a 
diagram such as Fig. 53. In this case the values of regular 
or specular reflection are for perpendicular incidence. The 
amount of light which a glossy varnished surface reflects 
is independent of the diffuse reflection of the coating be- 
neath this transparent coating. This regularly reflected 
light amounts to about four per cent for glossy varnish for 
perpendicular incidence. In Fig. 53, surfaces which reflect 
light only diffusely lie on the base line of the diagram, a 
perfect (theoretical) black being at zero and a perfect (the- 
oretical) white being at the right-hand corner. A matte 
commercial black is found at B and matte whites at D, E, 
and H. A glossy black paint is found at A reflecting about 
four per cent specularly and about four per cent diffusely. 
A glossy gray reflecting 50 per cent diffusely is found at C. 
Perfectly polished metals are found along the left-hand ver- 
tical line with average silver mirrors at K. A perfect mirror 
would lie at the upper corner and perfect reflectors (sur- 
faces reflecting all the incident light) would lie on the diag- 
onal, depending upon the percentage of specular and of 
diffuse reflection. Surfaces ordinarily encountered are 
found in the shaded area of the diagram. 
It should be noted that the values for specular or regular 
reflection in Fig. 53 are for perpendicularly incident light. 
The specular reflection-factor increases with the angle of 
incidence as shown in Fig. 10. A study of Fig. 53 will give 
one a vivid picture of the reflecting characteristics of most 
surfaces, those relatively few which exhibit spread reflection 
(such as metallic paints) being the only exception from a 
practical viewpoint. 
Reflection-Factors 
The measurement of reflection-factor of a surface is a 
difficult matter if high accuracy is desired because, in gen- THE LIGHTING VALUE OF PAINT 
233 
eral, the total incident light and the total reflected light 
must be measured by some integrating method. In the 
special case of a colorless surface which perfectly diffuses 
light, it is possible to measure its brightness with that of 
a standard diffusing surface of known reflection-factor re- 
ceiving the same illumination. A block of pure magnesium 
carbonate is a satisfactory standard for this purpose. Its 
reflection-factor is found to be as great as 98 per cent, this 
value being safe for a clean surface of pure material. 
High-grade commercial white paints have reflection-fac- 
tors between 75 and 80 per cent, although many white paints 
reflect only 70 to 75 per cent. Occasionally a white paint 
will be found to have a reflection-factor above 80 per cent 
but seldom is a value of 85 per cent found in paints. Com- 
mercial white paints differ much more in reflection-factor 
than is commonly suspected by those who merely casually 
examine them. Other things being equal, an 80 per cent 
white has an appreciable advantage over a 70 per cent white 
paint from a lighting standpoint where the systems of light- 
ing are such as to emit considerable light to the painted 
areas. 
The reflection-factor of paints changes appreciably even 
when kept in a clean dark place. White paints generally 
decrease about 10 per cent in reflection-factor in a year or 
more. Most paints lose in reflecting value although some 
colored ones may fade and thereby increase slightly. This 
change is of little interest in general practice in factories, 
offices, etc. because deterioration due to dirt is very much 
more rapid. From a practical viewpoint hiding power, 
specific gravity, cost, and character of the surface are im- 
portant factors. 
The measurement of reflection-factors of colored sur- 
faces is a very difficult matter because there are the addi- 
tional difficulties of color-photometry. It would serve no 
useful purpose to present a table of reflection-factors of  LIGHT AND WORK 
234 
surfaces in general because of the impossibility of identify- 
ing these various surfaces. Even dry powdered pigments 
used in the manufacture of paints are by no means constant 
in reflection-factor. The product of one factory may vary 
from time to time and may be quite different from the same 
pigment produced by another factory. Nevertheless, there 
is some value to such data and, therefore, in Table XVIII 
is presented the reflection-factors of dry powdered pigments. 
They give a fair idea of the reflection-factors of colored 
media and also show that the reflection-factor of a colored 
surface depends upon the spectral character of the illumi- 
nant. Reds, browns and yellows reflect less sunlight and 
skylight than the yellowish artificial lights. Conversely the 
greens and blues reflect less yellowish artificial light than 
white sunlight or bluish skylight. The illuminants are 
those commonly met in lighting practice. For the sake of 
accurate specification the lumens per watt and the filament 
absolute temperatures (in degrees Kelvin) are also given 
herewith. The tungsten gas-filled lamp operated at about 
15 lumens per watt and the filament temperature was 2780 
degrees Kelvin. The tungsten vacuum lamp operated at 8 
lumens per watt and its filament temperature was 2375 
degrees Kelvin. The flame-tint light is that corresponding 
to the light from a candle flame. For the sake of specifica- 
tion it is that light which would be produced by a tungsten 
filament operating at a low wattage so that its temperature 
was 2000 degrees Kelvin. 
Obviously the best manner of presenting data pertaining 
to reflection-factors is to accompany the data with the ac- 
tual specimens. This is done in the color-chart which is 
found herewith. This chart first appeared in the excellent 
bulletin by E. A. Anderson.41 The measurements of reflec- 
tion-factors of the present chart were made by A. H. Taylor 
in the most approved manner by means of the Taylor re- 
flectometer. Inevitably there is some slight variation intro- Reflection Factors 
The proportion of light reflected by walls and ceilings of various colors, that 
is, their Reflection Factors, have an important bearing on both the natural and 
the artificial lighting. The proportion reflected will depend somewhat upon the 
color of the incident light. The figures here given show what proportion of 
No. 1 
White 
Paper 
84% 
No. 9 
Ivory 
White 
77% 
No. 2 
Gray 
70% 
No. 10 
Caen 
Stone 
68% 
No. 3 
Gray 
69% 
No. 11 
Ivory 
73% 
No. 4 
Grav 
55% 
No. 12 
Ivory 
Tan 
56% 
No. 5 
Gray 
42% 
No. 13 
Primrose 
70% 
No. 6 
French 
Gray 
39% 
No. 14 
Lichen 
Gray 
69% 
No. 7 
Gray 
28% 
No. 15 
Pearl 
Gray 
73% 
No. 8 
Gray 
20% 
No. 16 
Silver Gray 
and Caen 
Stone 
51% of Colored Surfaces 
the light of Mazda lamps these painted surfaces reflect. Reflection Factors are 
of special usefulness in determining the Coefficient of Utilization (ratio of light 
delivered at the work to total light of lamps) applicable to an interior. The 
Reflection Factor of any colored surface can be approximated by comparing it 
with these samples. 
No. 17 
Buff Stone 
and Pale 
Azure 
37% 
No. 25 
Forest 
Green 
16% 
No. 18 
Buff 
60% 
No. 26 
Olive 
Green 
21% 
No. 27 
Pale Azure 
and White 
52% 
No. 19 
Buff Stone 
47% 
No. 20 
Tan 
28% 
No. 28 
Pale Azure 
39% 
No. 21 
Cocoanut 
Brown 
16% 
No. 29 
Sky Blue 
36% 
No. 22 
Satin 
Green 
63% 
No. 30 
Shell Pink 
51% 
No. 23 
Bright Sage 
and Ivory 
Tan 
43% 
No. 31 
Pink 
50% 
No. 24 
Bright Sage 
38% 
No. 32 
Cardinal 
Red 
16% THE LIGHTING VALUE OF PAINT 
235 
TABLE XVIII 
Blue 
Sky 
Noon 
Sun 
Tungsten 
(Daylight) 
Lamp 
Tungsten 
(Gas-filled) 
Lamp 
Tungsten 
(Vacuum) 
Lamp 
Flame- 
tint 
Light 
American Vermilion 
ii.7 
13-7 
14.8 
16.9 
17-7 
20.1 
Venetian Red 
o. 5 
10.6 
I! . 3 
12 • 5 
13.1 
14. 2 
Tuscan Red 
IO. I 
10.7 
II. I 
11.8 
12. 
12.8 
Indian Red 
9- 2 
9- 9 
10.3 
II. 
II. 2 
12. 
Burnt Sienna 
o. 1 
10.6 
II. 2 
12.2 
12.7 
13.5 
Raw Sienna 
30.3 
32.6 
34. 
35-7 
36.6 
38. 
Golden Ochre 
54.8 
58.1 
60. 
62.3 
63.4 
65.2 
Chrome Yellow Ochre.... 
28.9 
33- 
35-2 
38.7 
40.4 
43-5 
Yellow Ochre 
46. 
48.8 
5°. 5 
52.5 
53-4 
55 • 
Chrome Yellow (Medium) 
49-6 
54-5 
57-7 
61.3 
63- 
66. 
Chrome Yellow (Light)... 
70. 
76.5 
78.8 
81.1 
82. 
839 
Chrome Green (Light)... 
19. 
19-4 
18.5 
17.8 
17-5 
i7-3 
Chrome Green (Medium) 
14.2 
13-6 
13-2 
12.5 
12. 
11.6 
Cobalt Blue 
18.3 
16.2 
15.2 
14.1 
13. 
12.9 
Ultramarine Blue 
9-5 
7-4 
6.9 
6.1 
5-7 
5-2 
Reflection-Factors (in per cent) of Common Dry Powdered Pigments. 
duced in the preparation of the specimens but the reflection- 
factors as given are generally within one or two per cent 
of the standard specimens. The Sherwin-Williams Co. pre- 
pared the painted papers and the names of the paints are 
those used by that company. The chart was furnished by 
the National Lamp Works of the General Electric Co. 
Light diffusely reflected by a colored surface is tinted 
to some degree.19 When the ceiling receives a large per- 
centage of direct light from the lighting equipment, a 
conspicuous tint is given to the light reaching work-places 
if the ceiling is appreciably colored. This influence of 
colored surroundings is sometimes used to advantage but it 
is sometimes annoying as in coloring of daylight by adjacent 
colored walls. 
Colors for Walls and Ceiling 
For work-places, such as factories and offices, there are 
very definite color-schemes that cannot be surpassed in gen- 236 
LIGHT AND WORK 
eral satisfactoriness. For factories white ceilings and upper 
walls are best, in general, because they deteriorate rapidly 
enough. For the lower walls grayish green, olive green or 
dark buff is satisfactory. A lighter tint of the color used 
on the lower walls may be used on certain structural work 
or equipment, thereby giving the impression of a real deco- 
rative scheme. Workers have just as much right to look 
about them at pleasant surroundings. It is a growing prac- 
tice to paint different pipe-lines identifying colors. It seems 
that it would be quite sufficient to paint only the valves 
or elbows instead of trying to outdo the futurists. Black 
is often used for radiators and pipe-lines but there seems 
to be little justification for this depressing absorbing paint 
in most cases. The author avoids the use of pure white 
wherever possible. It is too often harsh, cold, uninviting. 
It is greatly tempered in this respect by a slight cream tint. 
It would be much better if the too prevalent white of hos- 
pitals, bathrooms and other interiors were tempered to 
ivory. 
In offices the ceiling which perhaps meets with most gen- 
eral satisfaction is a good white that has been tinted very 
slightly yellowish; in other words a cream tint. It seems 
that for visual comfort the reflection-factor of the walls 
should not exceed 50 per cent. A safe range is between 35 
and 50 per cent. Buff or buff stone in which the color is 
very much subdued by gray is generally very satisfactory. 
A medium grayish green is another satisfactory color which 
the author would rank second to the buff or buff stone. In 
rooms with high ceilings it is well to carry the ceiling color 
down the walls making an appropriate frieze of this ceiling 
paint. 
The matter of exposure is sometimes important. The buff 
or warm colors should be used in rooms of northern exposure 
in preference to the gray green. However, for southern 
exposure buff is still satisfactory, although many prefer THE LIGHTING VALUE OF PAINT 
the cooler color, such as gray green. The general superiority 
of the buff over the gray green is, in the author's opinion, 
enough to recommend its use when in doubt. 
There is also a psychological phase of color in factories. 
In the dull cold winter months the warmer color such as 
buff may have some advantages but undoubtedly the gray 
green or cooler colors are decidedly advantageous for summer 
months. Obviously we cannot have both, and therefore it 
appears best to decide on the green for all year round. In 
the factory generally the worker should have no difficulty 
in keeping warm. In fact, perspiration is associated with 
work. For this reason the cooler colors seem to be generally 
more desirable than the buffs. 
237 
Painted Machinery 
Throughout the entire industrial age manufacturers have 
persisted in painting machine tools and much other equip- 
ment either black or a very dark color closely akin to it. 
This not only contributed to the cheerlessness of poorly 
illuminated and ill-kept factories, but even in the modern fac- 
tory where the lighting is fairly well done and orderliness 
reigns, dark machine tools have a depressing effect. They 
could be painted somewhat lighter in color without showing 
grease, finger-marks, etc. Entirely aside from this we have 
the lighting value of paint in such places to consider. All 
machine tools and equipment should not be painted white. 
Such a condition would be insufferable. However, if they 
were painted a very subdued color of a reflection-factor 
of 20 to 25 per cent, much light would be saved by reflec- 
tion. But there is still an important phase of the use of 
paint and that is on surfaces near the work. The face- 
plate of a lathe, certain areas of the milling machine, the 
face of a press and many other areas near the tool or seat 
of operations can be painted a washable white or light gray. 
By reflection much light can be thrown upon the work and 238 
LIGHT AND WORK 
particularly in the shadows. On every hand we find such 
opportunities for using the lighting value of paint. No 
simple rules can be given. Each case is one involving light- 
ing and vision. Sometimes these painted areas would be 
excellent aids to vision as backgrounds. In other cases the 
advantages of light reflection might be offset by the de- 
sirability of a dark background. All these factors should 
be considered but it is certain that there are many applica- 
tions, small and large, of the lighting value of paints. 
Glare from a light-source or from the reflected image 
generally is due partially at least to an excessive brightness- 
contrast. The high-light on a painted surface - the re- 
flected image of a bright light-source - is more glaring 
when the surface is of low reflection-factor than when it is 
of a high reflection-factor. The brightness of the specularly 
reflected image is a matter of surface reflection and is not 
affected by the diffuse reflection. This accounts for the con- 
spicuous high-lights on glossy dark surfaces and the much 
less noticeable high-lights on glossy light surfaces. This is 
an advantage of light colored surfaces. 
Reflection of Invisible Radiation 
Although the reflection characteristics of paint and other 
media for ultraviolet and infra-red radiations are of no direct 
importance in lighting, they are of special interest occa- 
sionally. It has been known for years that zinc oxide 
photographed darker than other white pigments. This is 
because it absorbs ultraviolet radiation shorter than about 
366 mp in wavelength. Pfund has confirmed this and has 
shown that titanox, lithopone and sublimated white lead 
reflect only slightly radiations of shorter wavelength than 
320 mp. Magnesium oxide reflects practically all ultra- 
violet radiations at least as short as 250 mp in wavelength. 
In the infra-red region paints also differ in reflecting 
characteristics although there has been little systematic THE LIGHTING VALUE OF PAINT 
239 
investigating in this region. From the viewpoint of painting 
exteriors in hot climates at least this is important. In gen- 
eral, white exteriors which reflect much of the solar radia- 
tion are better than the darker ones which absorb much of 
the incident energy. It is easy to investigate this total 
absorption of energy by painting the exteriors of enclosures 
which are of the same shape and construction. By sub- 
jecting them to the same exposure of radiation and noting 
the temperature rise, the relative absorption can be ascer- 
tained. Even if metal strips are painted different colors 
and placed in the sunlight on snow, different rates of sinking 
into the snow will be noted. This gives a qualitative com- 
parison at least. Chapter XIII 
THE MOST EFFECTIVE INTENSITY OF 
ILLUMINATION 
IN other chapters the influences of lighting on vision and 
on human activity have been discussed at some length 
but no attempt was made to consider what is the most effec- 
tive lighting. In the present chapter data of other chapters 
and additional data are used in an attempt to establish more 
or less definitely the illumination intensity which appears 
to be most generally suitable for the most effective vision 
and the most efficient work. In these discussions the matter 
of proper lighting will be subordinated to that of adequate 
lighting, for it is assumed that the lighting system will be 
free from glare. That this is possible even at very high 
illumination intensities is demonstrated outdoors in the 
daytime. In fact, glare and lowered visibility are less easily 
eliminated at low levels of illumination than at high ones. 
In other chapters there is plenty of testimony in favor of 
much higher intensities of illumination than are commonly 
encountered indoors. We also know by experience that it 
is not comfortable to read black print on white paper in 
the direct sunlight at noon, although we can discriminate 
details on surfaces of lower reflection-factors without dis- 
comfort. Therefore, we know that the most generally 
desirable level of brightness lies somewhere between that 
due to an illumination of our surroundings of a few foot- 
candles and the 10,000 foot-candles at noon on a day in 
midsummer. 
Foot-candles, Reflection-Factors and Brightness 
The lighting specialist and others interested directly in 
lighting think and act largely in terms of foot-candles while 
240 MOST EFFECTIVE INTENSITY OF ILLUMINATION 
241 
those interested more particularly in vision should specify 
brightness. This is quite as it should be. The lighting 
specialist is not interested in any particular surface of a cer- 
tain reflection-factor but in all surfaces of the whole range 
of reflection-factors. Those who are concerned directly with 
vision naturally must deal first with the stimulus which is 
the brightness regardless of foot-candles and of reflection- 
factors. Nevertheless, the data must be transformable into 
both states. Particularly, those interested in lighting must 
visualize the brightnesses involved if they are to profit by 
much of the data pertaining to vision. 
Anyone, if he stops to think, realizes that illumination 
intensity is cause and brightness is effect. The link between 
the two is the diffuse reflection-factor of the surface illu- 
minated. It is customary to omit the word " diffuse " be- 
cause specular reflection applies only to polished or glossy 
surfaces and the brightness due to it is dependent upon the 
brightness of the light-source and not upon intensity of illu- 
mination. This point is brought out in Chapter XII in 
connection with Figs. 52 and 53. 
Owing to the fact that both brightness and illumination 
intensity are of interest in lighting and vision and that those 
interested more directly in lighting must be concerned 
primarily with foot-candles, the author devised the plan of 
presenting much of the data in the previous chapters in 
terms of foot-candles and diffuse reflection-factors. From 
the viewpoint of practice, so-called white is well represented 
by 80 per cent reflection-factor. By using this for the lower 
scale in many of the illustrations in Chapters VIII and IX, 
the data are readily visualizable. For convenience the upper 
scale in many of these illustrations represents foot-candles 
on a dark gray surface of 8 per cent reflection-factor. It 
is seen that these scales actually represent brightnesses, 
but the foot-candle unit still remains of primary importance. 
This proves to be a very practicable plan for the purpose 242 
LIGHT AND WORK 
of this book. It would be much simpler to use a brightness 
unit but the data would not be so easily visualized. 
In Fig. 54 is shown the complete relation between bright- 
ness, reflection-factor, and intensity of illumination. On the 
left-hand scale are found the foot-candles required in illu- 
minating surfaces of various diffuse reflection-factors (lower 
scale) to produce a brightness of one millilambert. If it is 
DIEEU5E REELECTION-EACTOR IN PERCENT 
FOOT-CANDLES PRODUCING A BRIGHTNESS OF ONE MILLILAMBERT 
(USE UPPER reflection-factor scaleJ 
FOOT-CANDLES PRODUCING A BRIGHTNESS OF ONE MILLILAMBERT 
01 Z3456789 IC> 
10 100 
9 90 
8 80 
7 70 
6 60 
5 so 
4 40 
3 30 
2 20 
I 10 
0 |0 20 30 40 50 60 70 80 90 100° 
Use left-hand scale with lowerscale 
Use right-hand scale with upper scale 
DIFFUSE REFLECTION-FACTOR IN PER CENT 
(use foot-candle scale on left) 
Fig. 54. Showing the illumination intensities in foot-candles for 
surfaces of various diffuse reflection-factors necessary to produce a 
brightness of one millilambert. The left-hand scale should be used 
with the lower scale for all values of reflection-factor greater than 
10 per cent. For smaller values the upper and right-hand scales should 
be used. 
desired to know the intensity of illumination for a surface 
of 20 per cent reflection-factor to produce a brightness of 
one millilambert, the curve directly above this value on the 
lower scale is seen to correspond to 4.64 foot-candles. For 
80 per cent reflection-factor it is one-fourth this value or 
1.16 foot-candles. For 10 per cent reflection-factor it is 
eight times the latter value or 9.28 foot-candles. For the MOST EFFECTIVE INTENSITY OF ILLUMINATION 
243 
sake of convenience the upper scale is for the range between 
zero and io per cent reflection-factor. The same curve 
applies to this range if the right-hand scale is used with the 
upper scale. For example, a brightness of one millilambert 
requires an illumination intensity of 46.4 foot-candles if 
the reflection-factor is only 2 per cent. It is obvious that 
from this diagram any other brightness can be considered 
by multiplying the results by its value in millilamberts. 
Furthermore, transformations can be made readily for 
brightnesses in terms of the other units presented in 
Chapter VIII. 
Fig. 54 vividly presents the very large range in foot- 
candles necessary to make ordinary surfaces, which differ 
widely in reflection-factor, appear equally bright. This has 
been so often overlooked in investigations and in presenting 
the results of them that it is emphasized here. For example, 
our so-called whites are generally used as backgrounds for 
visual experiments and the foot-candle scale is used in pre- 
senting the data without emphasizing that the foot-candle 
values apply to this background of particularly high reflec- 
tion-factor. As pointed out in a previous paragraph and in 
various illustrations in Chapters VIII and IX, the foot- 
candle values must be increased in proportion to the de- 
crease in reflection-factor. Everyday visual work, particu- 
larly in industries and partially in every activity, is often 
concerned not only with backgrounds of much lower reflec- 
tion-factor than so-called white, but the detail viewed against 
this background usually has a reflection-factor much greater 
than that of the so-called black. Furthermore, we are often 
obliged to see in shadow; that is, to discriminate details 
and differences in brightness and in color, on areas not 
receiving full illumination. Taking these facts into con- 
sideration, it is seen that the scale of intensities of illumina- 
tion (foot-candles) relating to data obtained with black on 
white in full illumination should be multiplied by a large 244 
LIGHT AND WORK 
" factor of safety " to make the data applicable to average 
visual activities. There are other phases which also tend to 
increase this " factor of safety." 
The relation between brightness, reflection-factor, and 
illumination may be expressed as follows: 
B = CIR 
where B = Brightness 
C = A Constant 
I = Illumination Intensity 
R = Reflection-Factor (diffuse) 
If I is expressed in foot-candles we have, 
Brightness in millilamberts = 1.076 IR 
Brightness in candles per sq. in. = 0.00221 IR 
For example, a surface diffusely reflecting 80 per cent 
of the incident light (0.8 reflection-factor) illuminated 
to an intensity of 50 foot-candles would have a bright- 
ness of,- 
1.076 X 50 X 0.8 millilamberts = 43.04 ml. 
or 0.00221 X 50 X 0.8 candles per sq. in. = 0.0884 
candles per sq. inch. 
Plotting Data Pertaining to Visual Functions 
In order to properly interpret data pertaining to visual 
functions it is necessary to recognize certain general charac- 
teristics of visual laws. Doubtless it has been noted that 
there is a general similarity among the curves expressing 
relations between brightness (illumination intensity multi- 
plied by reflection-factor) and visual acts such as acuity, 
sensitivity to brightness-difference, speed of discrimination, 
etc. Some of the illustrations which verify this are Figs. 
23, 26, 27, 40, 42, and 45. These and many other visual 
laws may be represented approximately by y = log x. 
From a scientific viewpoint it is more convenient and often MOST EFFECTIVE INTENSITY OF ILLUMINATION 
245 
more instructive to plot the data, yielded by investigations 
pertaining to vision, on a logarithmic scale. This has been 
avoided purposely in preceding chapters because the log- 
arithmic scales are confusing to those who are not used to 
thinking in terms of them. The logarithmic scale has an 
advantage in making it possible to plot a great range of 
brightnesses or illumination intensities in a small space and 
0 .01 .02 .05 .04 .05 .06 .07 .06 09 JO 
VALUES OF X fU5E Y SCALE TO RIGHT) 
-I 
2 
VALUES OF Y FOR X» 0 T0X - 0.1 
VALUES OF Y FOR X-OTOX'IOO 
I 
-2 
¥ = Log. X 
0 
-3 
-I- 
0 10 20 30 40 50 6 0 70 80 90 100 
VALUES OFX (USE Y SCALE TO LEFT) 
L4 
Fig. 55. The equation of the curve is y = log x and y is represented by 
the vertical scale on the left for corresponding values of x on the lower scale 
for a certain range of values (from x = o to x = 100). The identical dia- 
gram serves to represent another range of values (x = o to x = 0.1) as 
seen by applying the scale to the right and the upper scale. 
this is often desirable with visual data because of the ex- 
tensive range of sensitivity of the visual sense. This scheme 
was adopted in Figs. 22 and 30 because it was desirable to 
show the results over a great range. However, in most 
of the illustrations the regular scale sufficed to present the 
data in the range in which we are particularly interested. 
In Fig. 55, y = log x is plotted for two ranges of values 
and the same curve fits both of them. The left-hand scale 
for y values applies to the lower scale of x values between 
x = o and x=ioo. The right-hand scale for y values 246 
LIGHT AND WORK 
applies to the upper scale of x values between x = o and 
x = 0.1. Now supposing we had plotted the curve only 
for the range represented by the left-hand and lower scales. 
If this were an actual visual function (as it could be) and 
the lower scale represented foot-candles (for a specific 
background), we might conclude that the intensity of illu- 
mination should be greater than about 20 foot-candles in 
order to have the visual sense operating fairly efficiently. 
In other words, we might think that this value was above 
the " knee " of the curve and represented the beginning of 
a range of foot-candles where gain would not be appreciable 
for even great increases in illumination intensity. 
Let us lay aside these two scales for a moment and con- 
sider the upper and right-hand scales. (In fact, imagine 
the lower scale replaced by the upper one and the one on 
the left replaced by the one on the right.) The same curve 
fits these scales, but now, if we assume as before that the 
horizontal scale represents foot-candles, we might conclude 
that the intensity of illumination should be greater than 
0.02 foot-candles in order to have the visual sense operating 
fairly efficiently. This could be repeated for an infinite 
number of ranges of values with a different conclusion each 
time. Thus it is seen to be misleading to draw general con- 
clusions too freely from the form of a curve plotted for a 
single range of values. This mistake is often being made. 
With the foregoing limitations in mind, we find the form 
of the curve, and even the specific values applying to the 
region just above the " knee " of the curve, of considerable 
value. The form of the curve shows that the gain in y is 
less for a given increase in x as the value of x increases. 
Looking at it in any way, we find that when x is of a certain 
value it requires a greater increase in it to accomplish a 
certain increase in y than when x is of a lower value. If 
this curve represents a visual function such as sensitivity to 
brightness-difference we may state that an increase in sensi- MOST EFFECTIVE INTENSITY OF ILLUMINATION 
247 
tivity results from an increase in brightness (or illumination), 
but the rate of increase gradually lessens as the brightness 
increases. If the x values cover the entire range in which 
we are primarily interested, then further conclusions can 
be drawn with safety. 
In the relation of artificial lighting with production in 
various activities we are interested chiefly in the region 
from one to a thousand foot-candles. In fact, from the 
viewpoint of the intensities of illumination in general use 
indoors and including the relatively few of the higher in- 
tensity installations and even the highest intensities that 
might possibly be considered for general lighting at the 
present time, a range as high as too foot-candles would be 
ample. The data represented in Figs. 23, 24, 26, 27, 35, 
37, 38, 40, 42, 43, 45 and others apply to the range of foot- 
candles of interest to us at present in the consideration of 
effective lighting. For this reason we may use these curves 
for drawing conclusions as to the desirability of certain 
foot-candle values. 
A Resume of Data Bearing Upon Illumination Inten- 
sity and Vision 
In previous chapters we have seen that vision improves 
with increase of brightness or of illumination intensity. 
Now let us review these data with a view of establishing the 
minimum or perhaps a range of illumination intensity for 
the best vision. It should be borne in mind that these data 
were obtained in laboratory researches where there was 
little or no distraction, where the subject could be at ease, 
and the conditions in general were of the best. In all these 
researches excepting those actually involving " speed of 
vision," the subject had all the time necessary to arrive at 
a judgment. In other words, he did not feel hurried. Actual 
work conditions are quite different from the ideal condi- 
tions of the laboratory; therefore, it appears reasonable that 248 
LIGHT AND WORK 
a " factor of safety " should be allowed for work conditions. 
This is discussed later. 
The following is a summary of data bearing upon inten- 
sity of illumination for best vision as determined for various 
acts and test-objects. An attempt is made to estimate 
conservatively the minimum illumination intensity that is 
permissible if the eye is to function reasonably near its 
best efficiency. In those cases where the range of illumina- 
tion intensity extends to sufficiently high values the best 
range of illumination intensities is also indicated. In order 
to get a fairly complete view the values are given in terms 
of brightness in millilamberts and also in foot-candles for 
surfaces of 80 per cent and 8 per cent reflection-factors 
respectively. White light or the unsaturated yellowish light 
from ordinary light-sources and neutral or colorless surfaces 
are implied. 
The summary in Table XIX pertains chiefly to data 
already presented in Chapters VIII and IX. In only a few 
cases is it possible to give the value of brightness or illu- 
mination intensity where the best vision is attained. In 
only a few cases is it possible to give even the approximate 
range of more efficient vision. This is due to the fact that 
data are not available throughout a sufficient range. There- 
fore, in most cases only the approximate lower limit of the 
best range of brightnesses or illumination intensities is avail- 
able. It should be understood that visual efficiency con- 
tinues to increase above this lower limit, but that in the 
judgment of the author these are the lower limits for good 
vision even in the ideal conditions of the laboratory. In 
some cases visual efficiency is still very rapidly increasing 
at the highest brightness represented by the available data. 
Near the bottom of Table XIX certain values are taken 
from data by Nutting.44 What he terms discrimination- 
factor is the brightness of a field divided by the just 
noticeable brightness-difference in a portion of the field. MOST EFFECTIVE INTENSITY OF ILLUMINATION 
249 
TABLE XIX 
Data pertaining to the range of brightness for efficient vision. These 
are taken chiefly from the various illustrations indicated. 
Milli- 
lamberts 
(a) 
Foot-candles 
80% R 
(b) 
00 
Fig. 23. Best region for sensitivity to bright- 
ness difference is above 
9 
10 
100 
Fig. 26. Best region for discrimination of fine 
"black" details on a bright background is 
above 
9 
10 
100 
Fig. 37. Best region for rapid discrimination 
of "black" international test-object on a 
brighter background is above 
10 
12 
120 
Fig. 40. Best region for rapid discrimination 
of a small "black" dot on brighter back- 
ground is above 
87 
100 
1000 
Fig. 42. Best region for rapid reading of 
"black" print on a brighter background 
is above 
17 
20 
200 
Fig. 45. Best region for discrimination of 
"black" bars on a brighter background 
is above 
9° 
100 
1000 
Fig. 45. Best region for discrimination of 
"black" bars on a brighter background 
(preceded and followed by confusion pat- 
terns) is above 
35 
40 
400 
Fig. 57. Ditto but lower contrast is above.. 
180 
200 
2000 
Fig. 22. Minimum brightness-difference is 
discerned at 
300 
35° 
3500 
Nutting's data. Instantaneous threshold is a 
minimum percentage of adaptation-field 
brightness at about 
300 
350 
3500 
Fig. 22. Best region for discerning bright- 
ness-difference 
9-2600 
10-3000 
100-30000 
Nutting's data. Best range for instantaneous 
threshold 
100-1000 
115-1150 
1150-11500 
Nutting's data. Best range of discrimination 
factor 
10-100 
12-115 
115-1150 
This represents the ability to discriminate relatively large 
details (not the finest at least) when there is no appreciable 
color-difference; that is, the discrimination is due to per- 250 
LIGHT AND WORK 
ception of brightness-difference in a field of brightness. For 
example, if we have a large field of a certain brightness, we 
shall be able to distinguish a spot slightly darker or brighter. 
The brightness-difference necessary to see the spot depends 
upon the brightness of the background. Nutting's discrimi- 
nation-factor which he computed for a large range of 
brightnesses is a maximum for a range from io to 100 
millilamberts. This range of brightnesses is produced by 
12 to 115 foot-candles on a surface of 80 per cent reflection- 
factor and by 115 to 1150 foot-candles on one of 8 per cent 
reflection-factor. 
The data on instantaneous threshold were taken from 
Nutting's publication 44 of work by Blanchard, Reeves and 
himself. They exposed the eyes to a field of a certain 
brightness, 60 cm. square and viewed at a distance of 35 cm. 
This field brightness was suddenly extinguished leaving a 
small square (0.3 cm on a side) in the center of the large 
field. The brightness of the small square was adjusted 
until it was just discernible immediately after the bright- 
ness of the large field was extinguished. This procedure 
was carried out over a very great range of field brightnesses 
(0.000001 ml. to 10,000 ml.) and the instantaneous threshold 
(least brightness of small square that was discernible at the 
instant the brightness of the large field was extinguished) was 
determined throughout the entire range of brightnesses of the 
large field. For the range of field brightnesses from 100 to 
1000 ml., the instantaneous threshold brightness was of the 
order of magnitude of 0.2 per cent of the field brightness. 
This percentage rapidly increased as the field brightness 
decreased below or increased above this range. This is of 
interest in every-day vision because vision is being directed 
from one surface to another of different brightness. Speed 
of discrimination will in general be best for those ranges of 
brightnesses (levels of illumination) at which the eye can 
instantaneously discern the smallest fraction of the bright- 
ness to which it was just previously adapted. It is seen MOST EFFECTIVE INTENSITY OF ILLUMINATION 
251 
that this best range of brightness corresponds to a range of 
115 to 1150 foot-candles on a surface of 80 per cent reflec- 
tion-factor or of 1150 to 11500 foot-candles on one of 8 
per cent reflection-factor. 
These data on instantaneous threshold are so important 
that they are reproduced for a portion of the range of 
brightnesses in Table XX. The instantaneous threshold 
brightness has also been computed in per cent of the field 
brightness. Values of discrimination-factor, as defined by 
Nutting, and of minimum perceptible brightness-difference, 
have also been included in this table. The values in paren- 
theses are doubtless computed or estimated. 
TABLE XX 
Field 
Brightness * 
Millilamberts 
Instantaneous 
Threshold 
Brightness 
Brightness of 
Instantaneous 
Threshold in 
Per Cent 
of Field 
Brightness 
Minimum 
Perceptible 
Brightness 
Difference 
Discrimination 
Factor 
B/dB 
o.OOOOOI 
0.00000093 
93-o 
(1.00) 
1.0 
0.0000I 
0.0000042 
42.0 
(0.66) 
i-5 
O.OOOI 
0.000019 
19.0 
0-395 
2.5 
o. OOI 
0.000087 
8-7 
0.204 
4-5 
O.OI 
0.00039 
3-9 
0.078 
12.8 
O. I 
0.00175 
i-75 
0.037 
27.0 
i .0 
0.0082 
0.82 
0.021 
48.2 
10.0 
0.036 
0.36 
0.018 
57-5 
100.0 
0.191 
0.19 
0.017 
58.1 
1000.0 
2.14 
0. 21 
0.024 
4i-7 
10000.0 
(232.0) 
(2.32) 
(0.048) 
(20.9) 
Visual Sensitometric Data. 
* To transform these brightnesses into foot-candles on surfaces of 80 per cent 
and 8 per cent reflection-factor multiply millilambert values by 1.16 and n.6 
respectively. 
The Best Illumination Intensity Regardless of Cost 
Although in some aspects of lighting and vision we do 
not have adequate data available, in general there are  LIGHT AND WORK 
252 
enough data to warrant some conclusions. At least we may 
establish the best range of illumination intensities for gen- 
eral visual activity. These limits are not clearly defined but 
we do have sufficient data upon which to base a lower limit. 
The upper one we may safely conclude is so high as to be 
quite beyond hope of realization in artificial lighting at the 
present time. Even though it might be proved to be eco- 
nomically profitable to provide illumination intensities of 
several hundred or a thousand foot-candles of artificial light, 
the inertia of the user of light would not likely be overcome 
very easily. From the viewpoint of lighting practice the 
value of better lighting is of primary interest. We know 
that present conditions of artificial lighting can be greatly 
improved at a projit because of the increased safety, pro- 
duction, and cheerfulness. A question naturally arises as 
to where the law of diminishing returns may establish a 
dead line. This is discussed later. For the present let us 
consider the matter of illumination intensity quite aside from 
cost. 
In Table XIX we have seen a summary of the influence of 
illumination intensity upon vision and the illumination in- 
tensities of moment in the judgment of the author have been 
recorded. Of course, for any specific kind of work we could 
determine the important elements of vision that are in- 
volved and to what extent they play a part in the entire 
human activity. We shall find that brightness discrimina- 
tion is always essential. Sometimes visual acuity or the 
discrimination of the finest detail is involved. Nearly 
always fairly small details must be discriminated. In nearly 
all cases color-discrimination is involved but usually in a 
secondary, though helpful, manner. Relatively few areas 
are actually colorless and often there is a helpful color-dif- 
ference superposed upon brightness-difference. Colored 
backgrounds and colored light can be helpful or harmful in 
special cases. However, for the present it is best not to in- MOST EFFECTIVE INTENSITY OF ILLUMINATION 
253 
elude a consideration of color. Quality of light has been 
discussed in previous chapters, therefore it is not considered 
here. White light or the so-called white (yellowish) light 
of ordinary continuous spectrum illuminants was employed 
in obtaining the data under discussion. This is the kind 
of light that will be generally employed unless we find that 
we can safely improve upon Nature in this respect. 
Early in this chapter, it was pointed out that many visual 
functions are approximately logarithmic in character and 
therefore care must be exercised in drawing conclusions 
from the curves of plotted data. The form of these curves 
is similar to many efficiency curves of devices such as 
electric motors. If the range of values is that in which 
we are ordinarily interested it is of interest to draw such 
conclusions as are summarized in Table XIX bearing in 
mind the limitations. Column a gives the lower brightness 
at which vision begins to operate fairly efficiently. Inas- 
much as it is desirable to think in terms of foot-candles let 
us consider column b. Here it is seen that io foot-candles 
(on a surface of 80 per cent reflection-factor) is the lower 
limit of illumination intensity which the data indicate as 
desirable according to the judgment of the author. It 
should be noted that vision becomes better or more efficient 
for a great range of foot-candles above this " lower limit." 
This is indicated in all the illustrations mentioned in Table 
XIX and by the data in Table XX. The highest value of 
this lower limit is uncertain because most of the data avail- 
able do not extend far enough. However, the highest value 
(that the available data yield) of what seems a conserva- 
tive lower limit is about 100 foot-candles for 80 per cent 
reflection-factor or 1000 foot-candles for 8 per cent 
reflection-factor. 
Where the data are extensive enough we find the best 
range to be from io to 3000 foot-candles (for 80% R) or 100 
to 30,000 foot-candles (for 8% R). These data also yield  LIGHT AND WORK 
254 
the value for the most efficient vision as determined by the 
minimum perceptible brightness-difference and the instan- 
taneous threshold. It seem to be about 350 foot-candles 
(for 80% R) or 3500 foot-candles (for 8% R). 
TABLE XXI 
Summary of Values in Table XIX 
Foot-candles 
80% R 
8% R 
Lower limit of desirable intensities 
IO to IOO 
IOO to IOOO 
Range of desirable intensities 
io to 3000 
IOO to 30000 
Illumination intensity for best vision 
about 350 
about 3500 
Low Contrast Requires More Light 
In several chapters it has been pointed out that much 
of our laboratory work has been done with test-objects of 
high contrasts such as " black " details on ordinary " white " 
backgrounds. Conclusions drawn from these data under- 
estimate the importance of higher intensities of illumination. 
For this reason in many of the illustrations, particularly 
in Chapters VIII and IX, two scales of brightness have been 
supplied. These two scales are for foot-candles on surfaces 
of 80 and 8 per cent reflection-factor respectively. The 
foot-candle values on the 8 per cent reflecting surface are 
io times those on the 80 per cent scale for equal brightness 
of background. The former scale in all these cases shows 
the great importance of higher intensities because most of 
the brightnesses or backgrounds which we encounter in 
general vision lie somewhere between the two extremes. 
The only data available for speed of vision for lower 
contrasts than the usual black-white is that of speed of 
reading in Figs. 42 and 43 and that obtained by Cobb, Figs. 
56 and 57. These show the relatively greater importance MOST EFFECTIVE INTENSITY OF ILLUMINATION 
of higher intensities of illumination for low contrast than 
for high contrast. However, the actual contrasts between 
the type and background in the reading tests were not 
measured. Furthermore, the curves in these two illustrations 
255 
FOOT- CANDLES 
0 20 40 GO &0 100 IZO 
(6*1® REFLECTION-FACTOR) 
20 
SPEED OF DISCRIMINATION. 
1 
15 
10 
Visual angle = 1.73 minutes 
5 
0 I 2 3 4 5 6 7 8 9 10 II 12 
FOOT-CANDLES 
8O°/o REFLECTION-FACTOR 
Fig. 56. Showing the influence of intensity of illumination (or 
brightness of background) on the speed of discrimination of a test- 
object consisting of bars 3 mm. wide separated by intervals of the 
same width and viewed at 19.7 feet. The test-object was similar to 
b, Fig. 44, and it was preceded and followed by confusion patterns 
similar to a and b (Fig. 44) respectively. The high contrast (100:5) 
consisted of the usual " black " on " white." The low contrast 
(100:73) consisted of light gray bars on "white." The data are the 
averages of 21 observers. 
are made equal at one point so that they cannot be com- 
pared one with the other from a viewpoint of absolute 
values. 
Recently Dr. P. W. Cobb has completed some experi- 
ments with 2i persons for the test-object illustrated in Fig. 
44, but for two extremely different contrasts. In one case 
the contrast between the white background and the black 
bars in b, Fig. 44, was 100 to 5. In the case of low con- 
trast, the contrast between the background and the bars 256 
LIGHT AND WORK 
was zoo to 73. The test-object b was preceded and followed 
by " confusion patterns " a and c respectively. The 
data are plotted in Figs. 56 and 57. The data represented 
by the two curves in Fig. 56 were obtained with the bars 
of the test-object b, Fig. 44, 3 mm. in width separated by 
white intervals of equal width. The test-object was viewed 
at 19.7 feet, so that the visual angle between the black bars 
FOOT - CANDLES 
(a'lo REFLECTION-FACTOR) 
20 
0 400 600 1200 1600 2000 2400 
i 
-r=2Htrasi- /flops' 
5PEED OF DISCRIMINATION 
15 
10 
Visual angle - 1.15minutes 
5 
0 20 4 0 60 80 100 IZ0 140 160 180 200 220 240 
FOOT-CANDLES 
(80°|o REFLECTION-FACTOR) 
Fig. 57. Data from an investigation similar to that whose results 
are plotted in Fig. 56. Same observers, apparatus, etc., except that 
the bars of test-object, (and intervals between) were 2 mm. wide. 
Note the relatively high intensities of illumination required to attain 
the same speeds of discrimination as in Fig. 56. 
was 1.73 minutes. The test-object used in obtaining the 
data presented in Fig. 57 had bars and intervals 2 mm. in 
width. Viewed at 19.7 feet the visual angle between the 
bars was 1.15 minutes. 
It is seen that the speed of discrimination for the high- 
contrast test-objects increased very rapidly with brightness 
of the background (foot-candles on a surface of 80 per cent 
reflection-factor). For equal speeds of discrimination the 
low-contrast test-objects required very much higher inten- MOST EFFECTIVE INTENSITY OF ILLUMINATION 
257 
sities of illumination than the high-contrast ones. Further- 
more, it is noted that the foot-candle scale for Fig. 57 covers 
20 times the range of that for Fig. 56. In other words, a 
decrease in size of test-object from 1.73 minutes to 1.15 
minutes requires about 20 times the illumination intensity 
(background brightness) for equal speeds of discrimination 
of the two low-contrast test-objects. If the background 
had a reflection-factor of 8 per cent instead of 80 per cent 
the upper foot-candle scales would apply in Figs. 56 and 
57. It will be noted that in the latter case the foot-candles 
are very high values. 
Average Reflection-Factor 
Throughout these chapters brightness and illumination 
intensity have been related by expressing the brightness in 
terms of foot-candles on a surface of 80 per cent reflection- 
factor and also on a surface of 8 per cent reflection-factor. 
These represent extremes fairly well because the former is 
a so-called white and the latter is a very dark gray. If we 
were confronted all the time with surfaces of this very dark 
gray and were obliged to discriminate brightness-differences 
and details darker than the surface, we would need at least 
ten times more light than if the surfaces were so-called white. 
If we should choose 3000 foot-candles as the best intensity 
of illumination for the very dark gray surroundings, we 
would find this intensity too great for white surfaces. On 
the other hand, if the illumination intensity is barely suffi- 
cient for efficient discrimination of dark details on so-called 
white surfaces, certainly the intensity is insufficient for sur- 
faces of low reflection-factor and also for those in shadow. 
There is quite a range of illumination intensities, near 
the region of best vision, throughout which vision is fairly 
efficient. Taking this and other factors into consideration 
we may conclude that it would be fair to base our illumina- 
tion intensity on an average reflection-factor of 40 per cent. 258 
LIGHT AND WORK 
On consulting Table XIX it will be seen that a conservative 
and safe estimate of the good average brightness for effi- 
cient vision is 350 foot-candles on a surface of 80 per cent 
reflection-factor. Taking into account this average reflec- 
tion-factor of 40 per cent, we conclude that 700 foot-candles 
is approximately the best for work in which vision plays an 
appreciable part, provided the work receives full illumina- 
tion and the conditions are as generally satisfactory as in 
the laboratory. 
If we wish to provide for best vision in shadows receiv- 
ing only 10 per cent of the full illumination, the value of the 
illumination intensity outside the shadows would be 7000 
foot-candles. In Tables XIX and XXI it is seen that this 
is about twice the upper limit of desirable range for so- 
called white surfaces and is just at the upper limit which 
the data indicate as the desirable range for average reflec- 
tion-factor of 40 per cent. Therefore it is not desirable to 
provide fully for the best vision in shadows. Thus we ar- 
rive at a point where judgment alone is available and the 
matter may be considered as being involved in the factor 
of safety. 
Factor of Safety in Lighting 
Certainly it is not unreasonable to apply a factor of safety 
in lighting. If we could determine the best illumination 
intensity, what would be a reasonable factor of safety? 
Laboratory conditions are ideal compared with work con- 
ditions. The complication of movement of arms and body, 
the changing adaptation of the eyes to different brightnesses 
as we look from one surface to another, and various distrac- 
tions tend to reduce the efficacy of vision. It seems neces- 
sary to take these matters into account. Factors of safety 
in structural design have no necessary relation to factors 
of safety in lighting and vision but it is of interest to re- 
call that the former are often from 5 to 10. That is, it is MOST EFFECTIVE INTENSITY OF ILLUMINATION 
259 
not uncommon after making computations based on reliable 
physical data, to provide for strengths from 5 to io times 
that actually necessary for the purpose. 
We have a somewhat different situation in vision. Al- 
though vision becomes better or more efficient as the level 
of brightness (or illumination) increases, a level is finally 
reached beyond which ability to discriminate fine details 
and brightness-differences diminishes. Therefore, we must 
be careful not to provide for the best vision for backgrounds 
of low reflection-factor, for low-contrasts, and for details in 
shadow, at too great sacrifice of vision in the case of back- 
grounds of high reflection-factor which receive full illumi- 
nation. In Table XIX and XXI it is seen that the best 
range of intensity for so-called white surfaces extends to 
about 3000 foot-candles. Even beyond this there is a 
range where vision is quite satisfactory; however, it is well 
to bear this value in mind. It would be 6000 foot-candles 
for 40 per cent reflection-factor. 
In the preceding section it was concluded that 700 foot- 
candles represents a fair average value for the best vision 
when an average reflection-factor of 40 per cent is as- 
sumed. In a shadow receiving 10 per cent of full illumi- 
nation, this value would be increased to 7000 foot-candles. 
This leaves no room for increase by means of a factor of 
safety which would take care of the difference between 
vision in the laboratory and that under actual work con- 
ditions. As a matter of fact, if we should provide for best 
vision in shadows for average surfaces, we would find it 
necessary now to use a factor of safety less than unity in 
order to have efficient vision for so-called white back- 
grounds. Therefore, if we ignore vision in shadows we 
could have a factor of safety of about 4 because the upper 
limit of brightness for the best range of vision is 3000 foot- 
candles for a surface of 80 per cent reflection-factor and 
the best illumination intensity is about 700 foot-candles on 260 
LIGHT AND WORK 
a 40 per cent reflecting surface. Considering everything, 
this factor of safety may be too high because it is giving 
too much importance to vision in shadows. However, the 
advantage of being able to see well in shadows is experienced 
every day outdoors. 
The desirable value of factor of safety depends upon the 
illumination intensity which is actually under consideration. 
If we are dealing with one of 10 foot-candles and were con- 
sidering illumination intensity only from the basis of good 
vision, we could easily justify a factor of safety of 10 or 
even 30. 
This discussion is not introduced for the purpose of es- 
tablishing a fixed value of factor of safety but only for the 
purpose of suggesting the desirability of such a factor. 
Certainly under the prevailing intensities of illumination, 
factor of safety should receive consideration and at least 
it should be of a magnitude considerably larger than unity. 
To summarize again we may conclude that the following 
illumination intensities would be about the best as based 
on laboratory conditions. Certainly work conditions would 
not demand lower intensities. 
For reading black print on white background and not 
necessary to see in shadows, 350 foot-candles. 
For discriminating very dark details on backgrounds of 
20 per cent reflection-factor, 1400 foot-candles. 
For general work assuming an average reflection-factor 
of 40 per cent and seeing in shadows not necessary, 
700 foot-candles. 
For general work as in the preceding case but where see- 
ing in shadows is important at least 1000 foot-candles. 
The maximum upper limit of the desirable range of in- 
tensities is not far from 3000 foot-candles. Chapter XIV 
THE MOST ECONOMIC INTENSITY OF 
ILLUMINATION 
IF light did not cost anything, the most economic intensity 
of illumination would be that at which vision in general 
would be best. As seen in Chapter XIII, that intensity is 
in the neighborhood of 350 foot-candles for discriminating 
black details on white backgrounds, such as in reading. It 
would be somewhere between this and 3000 foot-candles 
for general work. The best intensity for average work 
conditions demanding the best vision is in the neighbor- 
hood of 700 to 1000 foot-candles. The average intensity 
of illumination indoors is less than one per cent of this 
latter value in artificial lighting and is not much more than 
one per cent of this value in congested commercial and in- 
dustrial districts in natural lighting indoors. 
In discussing the illumination intensities for best vision 
for various specific and average purposes, cost of lighting 
was purposely ignored. It is interesting to establish, ap- 
proximately at least, the ideal intensities for work. In 
practice it is to be expected that cost may be a limiting 
factor in striving to attain the ideal. If history repeats 
itself, the cost of artificial lighting will continue to decrease. 
If in the present century the cost of light decreases as much 
as it did during the past one, in the year 2000 light will 
cost only one thousandth of what it did in 1800. But we 
need not look that far ahead, because the cost of artificial 
light is low enough now to justify much better lighting than 
even the best installations supply today. Furthermore, the 
cost of light will continue to decrease. 
261 262 
LIGHT AND WORK 
Cost of Artificial Lighting 
The cost of lighting varies somewhat with location and, 
of course, from year to year. For this reason cost data are 
included here with some misgiving. Without such data cer- 
tain discussions cannot be presented, so it appears desir- 
able to present some data of this sort. E. W. Commery has 
very carefully computed the many items and checked his 
results with various contractors and others. These data 
are presented in condensed form in Table XXII and in 
several graphic charts. They apply to Cleveland in the 
winter of 1923-1924. The lighting system consists of 
porcelain-enamelled (RLM dome) steel reflectors and bowl- 
enamelled tungsten gas-filled lamps spaced uniformly on 
10-foot centers. The illumination intensity is uniform 
throughout the room which is 100 ft. by 200 ft., a repre- 
sentative unit of factory construction. The illumination 
intensities are not initial ones from clean lighting equipment 
and new lamps but are " average in service." That is, an 
average depreciation of 2 5 per cent in lighting is taken into 
account. The size of the lamp used is given in each case. 
The " materials " include everything with the exception of 
lamps; the latter are treated as a separate item. The in- 
stallation costs are separated from the operating costs per 
year. The computations are based on 2500 hours' operation 
per year. Interest, taxes, etc. are assumed to be 10 per cent 
and depreciation 12.5 per cent. The number of outlets re- 
mains constant for the entire range of illumination inten- 
sities. Electric energy being such a large item in operating 
cost at the present time and its cost being so variable, op- 
erating costs are given for 1 cent, 3 cents, and 5 cents 
per kw-hr. All items of cost in the table are given in 
dollars. MOST ECONOMIC INTENSITY OF ILLUMINATION 
263 
TABLE XXII 
Costs of installation and of operation per year (2500 hours) of a lighting 
system for a factory 100 ft. by 200 ft. for various intensities of uniform 
illumination. Porcelain enamel (RLM) steel reflectors and bowl-enamelled 
lamps are uniformly spaced on 10-foot centers. First-class wiring in conduits 
with adequate control of circuits. The foot-candles are average in service, 
allowance being made for an average depreciation of 25 per cent in initial 
light when equipment is clean and lamps are new. Materials include every- 
thing except the gas-filled tungsten lamps. These are in a separate item. 
Costs computed for Cleveland in 1924. Costs given in dollars. 
Foot-candles 
i-9 
3-8 
5-5 
9-2 
13 
21 
39 
61 
85 
128 
Watts per lamp 
Installation Costs: 
5° 
75 
100 
150 
200 
3°o 
500 
75° 
1000 
1500 
Materials 
$807 
$819 
$936 
$1058 
$1093 
$1499 
$1766 
$2104 
$2814 
$2814 
Labor 
269 
273 
312 
353 
364 
500 
589 
701 
938 
938 
Contractor 
285 
290 
300 
374 
750 
1030 
1214 
1445 
1932 
1932 
Total Installation 
1361 
1382 
1579 
1785 
1843 
2529 
2980 
3549 
4746 
4746 
Cost per outlet 
Operation Costs Per 
Year: 
6.80 
6.90 
7-9° 
8.90 
9.20 
12.65 
14-9° 
17-75 
23-80 
23.80 
Interest Taxes (10%).. 
$136 
$138 
$158 
$i79 
$184 
$253 
$298 
$355 
$475 
$475 
Depreciation (125%)... 
170 
173 
197 
223 
230 
316 
372 
444 
592 
592 
Lamp Renewals 
181 
221 
261 
321 
425 
657 
93° 
1442 
1520 
1920 
Maintenance 
300 
300 
3°o 
300 
300 
3°o 
3°o 
3°o 
3°o 
3°o 
Energy at 1 £ per kw-hr... 
Total Annual operating 
cost with electric 
energy at ifi per 
250 
375 
500 
75° 
1000 
1500 
2500 
375° 
5000 
75oo 
kw-hr 
1037 
1207 
1416 
1773 
2139 
3026 
4400 
6291 
7887 
10787 
per kw-hr 
1537 
1657 
2416 
3273 
4139 
6026 
9400 
10791 
17887 
25787 
per kw-hr 
2037 
2707 
34i6 
4773 
6x39 
9026 
14400 
21291 
27887 
40787 
In Fig. 58 the total annual operating cost in dollars is 
presented for the lighting installation for different intensi- 
ties of illumination delivered on the work-plane. Inasmuch 
as electric energy is such an important factor in operating 
cost, the total cost is given for three different rates, namely, 
1 cent, 3 cents, and 5 cents. It is seen that up to 130 foot- 
candles the curves are practically straight lines. 
In Fig. 59 it is seen that the total of fixed charges, such 
as interest and depreciation, is small, being for the most 
part less than 25 per cent of the cost of electric energy, or 264 
LIGHT AND WORK 
45000 
■40000 
TOTAL ANNUAL OPERATING COST IN DOLLARS 
35000 
30000 
25000 
20000 
15000 - 
10000 
5000 
0 10 20 50 40 50 60 70 60 90 JOO HO IZO 150 
FOOT-CANDLES 
Fig. 58. Cost of operation per year (2500 hours) of a lighting system 
yielding a uniform illumination on the work-plane for various intensities. The 
room is 100 ft. x 200 ft. RLM dome porcelain-enamelled reflectors, contain- 
ing bowl-enamelled tungsten lamps, are spaced uniformly on 10-foot centers. 
25000 
20000 - 
15000 
X/Qz 
\ M 
- 
\ \ 
\ 
DOLLARS 
10000 
5000 
Fix&d charges (interest,depreciation, etc.)_ 
0 10 20 30 40 50 GO 70 60 90 100 110 IZ0 130 
ILLUMINATION INTENSITIES - FOOT-CANDLES 
Fig. 59. Fixed charges, cost of electric energy at 3 cents per kilowatt hour, 
and total annual cost of operation for various uniform intensities of illu- 
mination on the work-plane. MOST ECONOMIC INTENSITY OF ILLUMINATION 
265 
about 20 per cent of the total annual operating cost based 
on 2500 hours' operation per year. It will be recalled that 
this is for a factory 100 feet by 200 feet. 
In Fig. 60 costs of lamp renewals, depreciation, and in- 
terest and insurance per year are presented. It should be 
noted that the scale of dollars to the left is very much dif- 
2000 
1500 
D0LLAR5 
1000 
Depreciotion (iZ'/t "I') 
ntereat etc. 
5000 
0 10 20 30 40 50 60 70 80 90 100 110 120 130 
Fig. 60. Lamp renewals, depreciation, interest and insurance for various 
uniform intensities of illumination on the work-plane. 
F00T-CANDLE5 
ferent from those of Figs. 58 and 59. The items represented 
in Fig. 60 are very small parts of the total as will be seen 
in Table XXII. 
In Fig. 61 the operating cost has been reduced to cents 
per foot-candle per square foot per year. Here again, owing 
to the relatively great importance and variability of the rate 
for electric energy, the data are presented for three different 
rates, namely 1 cent, 3 cents, and 5 cents per kw-hr. 
increased Production Can Purchase More Light 
As shown in Chapter X, an increase in production results 
from an increase in illumination intensity as far as any in- 266 
LIGHT AND WORK 
vestigations have extended; that is, to 20 or 25 foot-candles. 
Laboratory experiments, such as those discussed in Chapter 
IX, indicate that speed of vision continues to increase even 
beyond the limit to which the investigations have been 
extended, that is, beyond the region of 100 foot-candles for 
a surface or background of 80 per cent reflection-factor or 
to the region of 1000 foot-candles for a surface or back- 
ground of 8 per cent reflection-factor. Of course, work 
5 
4 
3 
Z 
I 
0 
CENTS PER FOOT-CANDLE PEP SQ.FT PER YEAR 
-MIoL-hour 
3*per kuu-hr 
1 * per km-hr. 
10 ZO 30 40 50 60 70 80 90 100 110 IZO 130 
FOOT-CANDLES 
Fig. 61. The operating costs of the lighting reduced to cents per foot- 
candle per square foot per year. Electric energy being such a large part of 
the total, the results are given for three rates, namely x cent, 3 cents, and 5 
cents per kw. hr. 
does not generally involve vision alone, but as long as vision 
is a part of the operation and it can be increased in speed, 
there will be an increase in production. It is reasonable to 
expect that illumination intensity will affect production in 
proportion to the ratio of the visual part to whole operation. 
It is also reasonable to expect that the character of the MOST ECONOMIC INTENSITY OF ILLUMINATION 
267 
visual part of the work will also have something to do with 
the effect of illumination intensity on production. 
In Fig. 49 we saw a gain in production of 12.5 per cent 
with a cost of lighting equivalent to only 2.4 per cent of 
the payroll when the intensity of illumination was 20 foot- 
candles. That is, the production under the new lighting 
system when it delivered 20 foot-candles to the work-plane, 
ZO 
ia 
16 
14 
12 
10 
8 
6 
4 
Z 
0 
PER CENT 
To 20 3 0 40 50 60 70 6 0 SO 100 110 
fOOT-CANDLES 
Fig. 62. Illustrating how much light can be purchased for actual increases 
in production. Data represented by A, C, and E are same as in Fig. 48. The 
Line BD extended both ways represents cost of light in terms of payroll 
(average wage .30 cents per hour). The line GH represents cost of light in 
terms of a doubled payroll (average wage 60 cents per hour). In this case, 
the production under the old lighting is taken as a basis. By projecting 
horizontally from E for example, the point F indicates the foot-candles (103) 
of the new lighting that could be purchased by the increase in production. 
was 12.5 greater than the production under the old light- 
ing system. The cost of lighting increases approximately 
uniformly with increase in intensity of illumination as seen 
in Fig. 58. With this in mind let us look at the results of 
Fig. 49 plotted on a different scale as in Fig. 62. The 
points A, C, and E represent the increases in production 
under the new lighting system for 6, 13, and 20 foot-candles 
respectively. The actual production under the original 
lighting (which was about 5 foot-candles of artificial and 
natural light mixed and poorly distributed) is taken as a 
basis. The increase-in-production curve ACE continues 268 
LIGHT AND WORK 
upward beyond E. Where it goes will be discussed later. 
The interesting point at the moment is how much light 
the increase in production can purchase. The line BDF 
represents the cost of light in per cent of the payroll for 
the actual wage-scale of 30 cents per hour. If a horizontal 
line is extended to the right from the point E, it intersects 
the line BDF at F. This means that the increased produc- 
tion under 20 foot-candles, if spent for light, would pur- 
chase 103 foot-candles of the new lighting. This is rather 
startling because it means that we can afford in this case, 
in many others, and perhaps in nearly all cases, intensities 
of illumination much greater than any actually in use for 
general lighting. It should also be noted that there will 
also be an increase in production in increasing the intensity 
of illumination from the 20 foot-candles to 103 foot-candles. 
Certainly all data point to such a result. By the same 
process it is seen that the line CD intersects BDF at D or 
at 66 foot-candles. In other words, the increase in pro- 
duction under 13 foot-candles can purchase 66 foot-candles 
of the new lighting. Likewise the increase in production 
A at 6 foot-candles would purchase 33 foot-candles of the 
new lighting. 
Now suppose the given number of workers received an 
average wage of 60 cents per hour instead of 30 cents. The 
cost of light in terms of this doubled payroll is only one- 
half (GH) as much as in the actual case (BF). Projecting 
the line EF until it meets GII, we would find that the inter- 
section would be at 206 foot-candles. In other words, with 
the higher-priced workers the increase in production under 
20 foot-candles would purchase 206 foot-candles of the new 
lighting. 
Now let us reconsider the data in Figs. 49 and 62 by 
eliminating the old system of lighting. This makes it pos- 
sible to see the results of increasing the intensity of illu- 
mination for a proper system of lighting without there MOST ECONOMIC INTENSITY OF ILLUMINATION 
269 
being any other variable such as a change from an old 
system to a new one. This is done in Fig. 63. Now the 
production under 6 foot-candles from the new system of 
lighting is taken as a basis and the percentages of increase 
in production are computed. At 13 foot-candles the in- 
crease was 3.8 per cent and at 20 foot-candles it was 8 
20 
18 
16 
14 
12 
10 
ft 
6 
4 
2 
0 10 20 30 40 50 60 70 80 30 100 110 
PER CENT 
Fig. 63. Similar to Fig. 62 excepting that the production under 6 foot- 
candles from the new lighting system (Fig. 48) is taken as a basis. This 
eliminates the matter of different systems of lighting. The cost of light in 
terms of the payroll is represented for several wage scales, 30 cents per hour 
being the actual one at the factory where Hess and Harrison obtained the 
data. 
FOOT-CANDLES 
per cent. Projecting the increase at 20 foot-candles A 
across horizontally it meets at B the line representing the 
cost of light in per cent of the payroll. This point B is at 67 
foot-candles when wages averaged 30 cents per hour. Thus 
it is seen that the increase in production (with the same 
lighting system) under 20 foot-candles as compared with 
that under 6 foot-candles will purchase 67 foot-candles of 
the new lighting. What further increase in production 
would result in passing from 20 to 67 foot-candles is dis- 
cussed later. Certainly it would be appreciable. The same 
process can be applied to any other increase of production. 
It is also interesting to see where the line AB would cut 
the lines representing the cost of light in terms of the pay- 
roll for other average wages per hour. For example, if the 270 
LIGHT AND WORK 
average wage were 45 cents per hour the line AB cuts this 
line at C. A summary of the results from Fig. 63 is pre- 
sented in Table XXIII. It is seen that as the payroll in- 
creases the amount of light that even a small increase in 
production can purchase is very great. For example, an 
increase in production of 2 per cent pays for 17 foot-candles 
when wages are 30 cents per hour and pays for 50 foot- 
candles when wages are 90 cents per hour, For the latter 
wage scale an increase in output of 8 per cent pays for 
200 foot-candles. 
TABLE XXIII 
Showing how much light can be purchased by the increase in production 
resulting from a given increase in intensity of illumination. The lighting system 
and the number of workers remaining constant but the illumination intensity 
and wage-scale being changed. 
Per Cent 
Increase in 
Production 
Actual 
Foot-candles 
Foot-candles that can be Purchased by the 
Increase in Production if Wages Per Hour are: 
30 cents 
45 cents 
60 cents 
90 cents 
o 
6.o 
2 
9-4 
17 
25 
33 
5° 
4 
130 
33 
5o 
67 
100 
6 
16.6 
5o 
75 
100 
150 
8 
20.0 
67 
100 
133 
200 
It simplifies matters to adhere to one set of data in pre- 
senting these various analyses, although it is recognized that 
there may be criticism to the effect that the data obtained 
by Hess and Harrison 43 are stressed too much. Besides the 
excuse of simplicity in treatment, their data are the best 
available from factories at present and are quite complete 
and convincing. The investigation was carefully conducted 
over a period of ten weeks in an approved manner. Fur- 
thermore, their data are strongly supported by much 
laboratory data, some of it, such as reading, being actual 
production data. Various factory tests also yield similar MOST ECONOMIC INTENSITY OF ILLUMINATION 
271 
results. In fact, there is a gratifying consistency among all 
these results indicating that, if the illumination intensity 
were increased in offices and factories to 20 foot-candles, 
there would be an increase in production from 10 to 15 
per cent at a cost of lighting amounting to one to three 
per cent of the payroll. If this is true, an illumination in- 
tensity of 100 foot-candles is certainly practicable at the 
present time. 
In Chapter XIII it was shown that intensities from 350 
to 1000 foot-candles are ideal. Are such intensities prac- 
ticable at the present time considering the cost of lighting? 
Unfortunately we do not have production data extending 
far enough to state positively. Certainly we have seen that 
100 foot-candles are not to be considered far from practicable 
at least. 
Form of Curves Representing Increase in Production 
As has been pointed out before, many visual functions, 
including those of speed of discrimination, have similar 
forms. These are satisfied fairly well by logarithmic equa- 
tions. This would have been clear if the author had adopted 
the simple expedient of plotting much of the data presented, 
on logarithmic scales. This was purposely avoided because 
logarithmic scales are confusing to those unused to them 
and, furthermore, in most cases the range of brightness or 
of illumination in which we are chiefly interested can be 
taken care of by the usual scale. It would simplify and aid 
greatly the consideration of the economics of lighting, if 
equations relating illumination intensity and increase in 
production were available. To obtain these it is now neces- 
sary to perform actual tests for a given operation. Ob- 
viously, it involves a large amount of work and expense 
to carry this out even to 40 or 50 foot-candles. Laboratory 
work can be done very much simpler and under better con- 
trol, so it is likely that laboratories must continue to supply  LIGHT AND WORK 
272 
the data that encourage progressive men to increase illu- 
mination intensities much beyond those that have been 
used. Furthermore, there is an unwarranted attitude 
against laboratory work. Too often do we hear the so-called 
" practical " man speak of laboratory results as having little 
bearing upon practice. As a matter of fact they have a 
great deal of bearing and, furthermore, they are very de- 
pendable data in most cases and pertinent to activities in 
the office or factory. The fault lies largely in lack of 
interpretation of such data in terms of practice. 
The author has given considerable study to the forms of 
various curves relating vision and speed of vision to illu- 
mination intensity, and has found that certain simple log- 
arithmic equations represent most of these curves very well. 
L. L. Holladay has kindly gone over much of the data and 
checked the results. Of course, it is possible to develop 
equations that exactly fit the results. By plotting the data 
logarithmically, extrapolation beyond the range of avail- 
able data can be done without great risk of appreciable 
error if the extrapolation is not carried too far. This is be- 
cause a range from 20 to 200 foot-candles is logarithmically 
equal to a range from 2 to 20 foot-candles. 
In Fig. 64 are presented the simplest logarithmic equa- 
tions that satisfy certain representative data pertaining to 
speed of reading or of discrimination of various test-objects. 
The full lines represent the respective equations and the 
circles represent some of the original data to illustrate how 
closely even the relatively simple equations represent the 
actual results of speed of discrimination. On the right 
of each curve is found the number of the Figure in which 
the original data are plotted. The horizontal scale is that 
of foot-candles on a surface of 80 per cent reflection-factor 
and each curve has its own scale as indicated. The verti- 
cal scale is that of speed of vision, each curve having its 
own scale. It is seen that the agreement between the MOST ECONOMIC INTENSITY OF ILLUMINATION 
273 
plotted equation and the actual data is quite close. For 
various details the reader should refer to the original 
illustrations in Chapter IX. 
200 
\bO 
120 
80 
160 
IZO 
80 
40 
Ji- 
160 
120 
80 
40 
J2_ 
200 
160 
IZO 
60 
40 
0 
Speed =56 + 78 log (Foot-candles) oA Fig.4Z 
' • »-B Fiq.42 
Speed = 81+29 log (Foot- candles) 
10 20 
-Fig. 40 
Speed-20+ 63 log (Foot-candles) 
50 
100 
A Fig. 45 
-Speed = 95 log. (Foot-candles) 
-Speed =63t20 log (Foot-candles) 
50 100 
B Fig. 45 
rig. 37 
(3.451) 
SPEED OF VISION 
Speed= 13.5 (foot-candles) (Visual angle -0.5) 
(in minutes) 
5 
10 
FOOT-CANDLES 
Fig. 64. The simplest equations which fairly well represent the data pre- 
sented in certain illustrations in Chapter IX have been determined. They are 
represented by the curves. To show how well they represent the actual data, 
some of the latter are indicated by the circles. 274 
LIGHT AND WORK 
A Brief Summary of Fig. 64. 
Curve A, Fig. 42. The relative speed of reading black print on the 
gray background is fairly well represented by 
s = 56 + 78 log I 
Where 5 is relative speed of reading and I is the intensity of illu- 
mination in foot-candles. The data are almost perfectly represented by 
Curve B, Fig. 42. The relative speed of reading black print on a 
white background is fairly well represented by 
5=664-71 log (Z - 0.8) 
S= 8i + 29 I 
The data are almost perfectly represented by 
S = 84 + 24 log (Z - 0.23) 
Curve Fig. 40. The speed of discrimination of a black dot on a 
white background is well represented by 
5= 20 + 63 log I 
Curve A, Fig. 45. The speed of discrimination of black bars in a 
white background is well represented (for values of 1 above 5 foot- 
candles) by 
5 = 95 log Z 
The data are represented better throughout its range up to 7 = 50 
by 
Log S = 1.57 + 0.38 log (Z - 0.5) 
Curve B, Fig. 45. The speed of discrimination of black bars on a 
white background, when the test-object is preceded and followed by 
confusion patterns, is well represented by 
5= 63 + 20 log I 
The data are also represented by 
Log S = 1.43 + o.2i log (Z- 0.5) 
Curve Fig. 37. The data on speed of discrimination of a split ring 
(international test-object) for four visual angles (1.25, 1.73, 2.49, 
and 3.45 minutes) is fairly well represented by 
Log S = 1.13 4- 1.5 log (7 - 0.5) + 0.5 log I MOST ECONOMIC INTENSITY OF ILLUMINATION 
275 
Where V is the visual angle in minutes of the missing portion of 
the ring. This equation for a test-object of 3.45 minutes visual angle 
is the one plotted in Fig. 64. 
Simple equations approximately representing the data are as follows 
for the four visual angles: 
For 1.15 minutes, S= 2.7 14.2 log I 
For 1.73 minutes, S- 12.5 j. 79 log I 
For 2.49 minutes, 5 = 42 77 log I 
For 3.45 minutes, 5 - 59 -j- 150 log I 
Many other examples besides those in Fig. 64 could be 
presented to illustrate that speed of vision and various 
visual functions are fairly well represented by logarithmic 
equations. This aids us in visualizing what is likely to be 
the effect on production if we should increase the intensity 
of illumination far beyond those actually used in production 
tests. Certainly our explorations in this realm in advance 
of actual tests are not solely speculative because we have 
the support of these various relationships. For this reason 
let us return again to the production tests presented in Figs. 
48, 62 and 63. The relation between actual production and 
intensity of illumination is fairly represented by the follow- 
ing two equations, 
P = + 65 log/ 
Log P= 2.577 + -064 log 1 
where P is the average number of pieces inspected per 
person per hour and I is the intensity of illumination in 
foot-candles. However, the per cent increase in pro- 
duction interests us most. Considering only the new 
lighting installation, and taking the production at 6 foot- 
candles as a basis, the increase in production G is related to 
the intensity of illumination I as shown in the equation 
G = 15.17 log I - 11.8 276 
LIGHT AND WORK 
This curve is plotted in Fig. 65, the portion AD and be- 
yond is that for which there are no test data. It has already 
been pointed out why we may have some confidence in 
extrapolating. The logarithmic plot yields a straight line 
and we can extrapolate to somewhat higher values of foot- 
candles with considerable confidence that the results are 
going to be roughly approximate at least. Most of the 
available data on visual functions support this procedure. 
24 
20 
16 
12 
8 
fl Lb 15 L4 hZ LS zJF'z.Z 2.4 77 
Log, toot-candles 
6F.C . 
lOf.C. - 
ZO EC- 
40 EC- 
ioo r.c- 
400 EC- 
A 
32 
23 
Z4 
20 
Per cent 
10 
16 
12 
a 
4 
0"E ZO 40 60 80 100 120 140 160 180 ZOO ZZO Z40 
D 
i v 
Jo> ♦ 
l\ 
i °uX _ 
1 A °v 
1 °\ y 
1 A 
r 
PER CENT 
F00T-CANDLE5 
Fig. 65. An equation was determined for the data on increase in pro- 
duction Fig. 48 and is represented by the curve EAD. The cost of light in 
per cent of payroll (average wage 30 cents per hour) is represented by BD 
extended both ways. EAD intersects BD at 160 foot-candles. The cost of 
light in per cent of payroll if average wage were 60 cents per hour is also 
represented. 
In the upper part of Fig. 65 is shown the method of 
obtaining the logarithmic equation which fits the data 
fairly well. The left-hand scale is per cent increase in 
production over that at 6 foot-candles. The logarithms of 
the foot-candles are plotted on the horizontal scale. As- 
suming that the relation between per cent increase in pro- MOST ECONOMIC INTENSITY OF ILLUMINATION 
277 
duction and the logarithm of the foot-candles is linear, we 
draw a straight line which seems to fit the data represented 
by circles. This straight line is then extended. A few 
points on a scale of actual foot-candles are shown at the 
top. This illustrates the fact that, assuming the linear re- 
lation just mentioned, we can extrapolate from 20 to 200 
foot-candles without extending the actual straight line un- 
reasonably far. Of course, this does not prove that the 
logarithmic equation accurately represents the results which 
would be obtained in the region unexplored by actual test. 
However, it is presented for what it may be worth and to 
give a picture of what might be expected. 
The relation between intensity of illumination and the 
cost of light in terms of the payroll (average wage 30 cents 
per hour) is represented by the line BD extended. Fig. 58 
supports this straight-line relation. It will be noted that 
the curve AD representing G, the per cent increase in pro- 
duction, intersects BD at D. The latter point is at 160 
foot-candles. In other words, if the equation fairly repre- 
sents the relation between increase in production and foot- 
candles, at 160 foot-candles the increase in production 
(over the production at 6 foot-candles) equals the cost of 
lighting. Supposing this factory had a well-designed light- 
ing system supplying 6 foot-candles to the work-plane. If 
the intensity of illumination were increased to 160 foot- 
candles (assuming accuracy of the equation relating G and 
/), this new lighting would be obtained at a net cost of 
nothing, production would have increased 22 per cent, and 
there would be the additional advantages of decreased 
spoilage and increased safety and general cheerfulness. If 
in this same factory the wage was 60 cents per hour the 
cost of light in per cent of the payroll would be represented 
by the straight line passing through the point C. This line 
extended would meet the curve AD extended in the 
neighborhood of 400 foot-candles. LIGHT AND WORK 
278 
In Fig. 63 and Table XXIII it is seen that the actual 
increase in production at 20 foot-candles would pay for 67 
foot-candles for the group of persons involved at the actual 
average wage of 30 cents per hour. It would pay for 133 
foot-candles if the wage scale were 60 cents per hour. 
These are facts directly arising from actual data. In Fig. 
65 it is indicated that if the curve relating foot-candles and 
per cent increase in production is approximately correct, 
the increase in production over that at 6 foot-candles would 
pay for 160 foot-candles when the wage is 30 cents per 
hour and would pay for about 400 foot-candles if the wage 
were 90 cents per hour. No claim to accuracy in prediction 
is made, although for this test in an inspection department 
where vision is an important part of the operation, it seems 
that we can have confidence in at least a portion of the 
curve between A and D. Certainly vision in general would 
improve up to 350 foot-candles at least, and in many cases 
up to higher intensities. In fact, it has been indicated that 
we can expect improvement in vision up to 1000 foot-candles, 
at least, for general factory work. 
It should be noted that the discussion of Fig. 65 pertains 
to increase in production due solely to increase in illumi- 
nation intensity from a well-designed lighting installation. 
The increase in production in changing from a poorly 
designed installation to a well-designed one is usually con- 
siderable, quite aside from the increase due to increase in 
intensity of illumination. Chapter XV 
VISIBILITY AND SAFETY 
IT has been amply proved in preceding chapters that ade- 
quate intensities of illumination are necessary for the 
best vision and therefore for efficient work. They also 
decrease spoilage, increase safety, decrease eye-fatigue, and 
tend to make work-places more cheerful. Illumination in- 
tensities may be adequate, but still the lighting may be im- 
proper. The light-sources may be glaring, polished surfaces 
may reflect bright images, shadows may be harsh, or the 
direction of light may not be satisfactory. All these may 
reduce the ability to see and may cause visual discomfort 
with all its complex ramifications. Improper lighting, with 
the exception of inadequate intensity, has not been discussed 
in preceding chapters. This relates more to the design and 
control of lighting which is beyond the scope of this book. 
Nevertheless, it appears desirable to touch upon some of 
these points, briefly at least, because adequate lighting loses 
much of its value if it is improper. 
The efficiency of a lighting system is not measured solely 
by the efficiencies of the light-source and the lighting equip- 
ment, and the reflection-factors of ceiling, walls, etc. It 
is not merely a matter of cents per foot-candle per square 
foot per year. Efficiency of lighting finally involves effi- 
ciency of vision and also eye-comfort. In other words, 
efficiency in lighting is the ratio of satisfactoriness to cost. 
It has been well proved that inadequate and improper 
lighting causes defective vision. In some industrial plants 
it has been found that as many as 50 per cent of the 
workers have defective vision. There is an increasing per- 
279 280 
LIGHT AND WORK 
centage having defective vision as the average age of a 
group increases. Much of this increase can be attributed 
to bad lighting. Adequate light properly shaded and dif- 
fused can do much to retard this increase in defective vision 
with age. Furthermore adequate lighting should not be 
based solely upon a consideration of perfectly normal eyes. 
A factor of safety should be allowed for that great 
percentage whose vision is defective. 
Glare 
This is a term which is used rather loosely. Perhaps it 
is best to confine it to a lighting condition which causes 
visual discomfort. Glare may be due to 
a. Excessive brightness of a light-source or of its 
image reflected from polished or mirror-like 
surfaces. The brightness that is not glaring 
depends upon the adaptation of the eye, the 
background against which it is seen, and some- 
what upon its size or visual angle. For general 
lighting conditions indoors the maximum bright- 
ness should not exceed 20 candles per square 
inch and it is best to keep it as far below this 
value as possible. This can be done by properly 
diffusing the light. 
b. Perhaps our most common source of glare is exces- 
sive contrast. Outdoors at night a flaming match 
may be glaring amid the very dark surroundings. 
The sky is not glaring out doors but a patch 
seen through a window amid the relatively dark 
surrounding walls is commonly somewhat trying 
on the eyes. If the backgrounds against or 
amid which lighting equipment is viewed are 
of fairly high reflection-factor, and if they re- 
ceive an appreciable amount of light, glare will VISIBILITY AND SAFETY 
281 
generally be eliminated if the lighting equip- 
ment is satisfactorily diffusing. 
c. A bright area of small visual angle may not be 
glaring but a large area of the same brightness 
may be very glaring. In other words, volume of 
light incident directly on the eye is a factor in 
glare. 
d. Light sources too close to the line of direct vision 
are more glaring than if they are further away 
from this line. They are also less glaring if 
they are in the extreme upper than if in the ex- 
treme lower part of the visual field. 
e. The light-sources may be properly shaded from the 
eye but mirrored or polished surfaces may 
reflect an image toward the eyes. Therefore 
such surfaces should be avoided wherever pos- 
sible, because they can reduce or even defeat the 
best efforts of the lighting specialist. 
/. The time of exposure to glaring conditions has 
something to do with the effectiveness of the 
glare. A bright light-source for instance can 
be looked at for a moment without as much glare 
resulting as if it were looked at for a few seconds. 
On the other hand if the source is not too bright 
the retina can become adapted to it if looked 
at directly. This does not mean that no harm 
results. In fact, this adaptibility is unfortunate 
because it makes persons careless about glare. 
These are the chief factors in glare. No attempt is made 
here to give limiting values.40 Anyone with some feeling, 
some ability to observe, and some common sense can readily 
learn to sense conditions discomforting to vision. 
It is convenient to divide glare into three classes.47 Veil- 
ing glare obscures details by a veil of light spreading from a LIGHT AND WORK 
282 
brilliant light-source near the line of vision, or more rarely 
by diffused light in smoky, foggy or dusty air. 
Dazzle glare is uncomfortable, distracting, and reduces 
the ability to see. This is the most common type of glare. 
It can be due to light-source or lighting unit or to the images 
of these reflected from polished surfaces. 
Blinding glare is due to an extreme brightness that greatly 
decreases for a time the ability of the retina to respond to 
light. Powerful after-images remain for some seconds or 
even a minute or two, depending upon the brightness of the 
glare source. 
All these types of glare decrease the ability to see and 
therefore not only decrease the efficiency of a worker but 
increase the possibility of accident. 
Brightness of Light-Sources and Lighting Equipment 
An idea of the brightnesses encountered is obtained from 
the data in Table XXIV. The brightnesses of clouds and 
various skies are presented in Table III. 
TABLE XXIV 
Noon sun .... 
Approximate Brightness 
in Candles per 
Square Inch. 
Crater of ordinary carbon arc 
Tungsten filament vacuum lamp (8 1. p. w.) 
Tungsten filaments of gas-filled lamps 
Tungsten filament of gas-filled go-watt lamp 
Tungsten filament of gas-filled goo-watt lamp .... 
Diffusing bulb of go-watt tungsten lamp 
Enamelled portion of bowl-enamel tungsten lamp ... 
Mercury vapor tube .... 
1060 
2400-gooo 
2630 
45oo 
8 
10-ig 
t e 
Brightest sunlit cloud . 2O 
::::::::::::::::::::::::::::::: a~4 
Globes of good diffusing glass of various 
diameters and containing a 100-watt lamp 
Diameter 
5 inches 
A < 
H 
6 
CC 
Q I 
Cl 
7 
cc 
d*x 
2 'Z 
cc 
8 
cc 
'O 
I 8 
Cl 
IO 
cc 
ic 
14 
(( VISIBILITY AND SAFETY 
283 
Recommended globe sizes of good diffusing glass 
Lamp size 
Globe diameter 
50-75 
8 
2.5 
75-100 
10 
2.5 
100-150 
12 
3-0 
150-200 
14 
3-5 
200-300 
16 
4-o 
300-500 
18 
5-o 
Eye-Strain 
This is the result of inadequate or of improper light. 
When light is inadequate the work must be held too close to 
the eyes. This results in a strain of the eye-muscles. Glar- 
ing light-sources tend to attract the attention and the eye- 
muscles must strain themselves to keep the eyes on the work. 
Glaring conditions cause one to " pucker his eye-brows " 
with a resulting strain. If the lighting is quite non-uniform 
there is a noticeable effect of the necessity for pupillary ad- 
justment in looking from one brightness to another of a 
different magnitude. Eye-strain is not conspicuous under 
the high intensities of daylight, and under proper conditions 
eye-strain can be eliminated in artificial lighting. 
Eye-Fatigue 
Under proper and adequate lighting, the eyes are not 
fatigued any more than the body is under proper working 
conditions. Everyone with reasonably normal eyes is quite 
satisfied with the high intensities of daylight. Rarely is a 
day so bright in summer that the eyes are not comfortable, 
although we have become used to shading the eyes by means 
of hats even from much of the sky. Eye-fatigue is much 
less under high intensities of proper lighting than under low 
intensities of lighting which are otherwise proper. There is 
some indication that light approaching daylight quality 
(white) is more desirable for long periods of serious work 
than yellowish or less natural light. Certainly eye-fatigue 
is more noticeable after long periods of reading or close 
work under a low intensity of illumination than under an 284 
LIGHT AND WORK 
adequate one. Eye-fatigue is perceptible when the sur- 
roundings are quite dark as compared with the work. Of 
course, the eyes are rested by looking from the bright work- 
surface to some area less bright but this is a different 
matter. 
Visibility 
A reduction in ability to see is one of the results of glare. 
Therefore, improper lighting conditions decrease the effi- 
ciency of the workman. Glaring light-sources or their 
reflected images reduce visibility very greatly if they are 
near the line of direct vision. Light is scattered in the eye 
and there are certain diffusion effects in the retina or some- 
where in the visual processes that reduce the ability to see. 
If a light-source is very bright, the after-image may be 
blinding for a long time. (See Fig. 34.) Visibility gradually 
decreases when the eyes are subjected to improper lighting 
conditions. 
Visibility of black details on a background of small area 
is best when the immediate surroundings are about as bright 
as the background of the test-objects. 
Safety and Accidents 
Improper and inadequate lighting is responsible for 
about 20 per cent of all traffic and industrial accidents.17 
R. E. Simpson 48 has stated that each year there are more 
accidents in this country due to inadequate and improper 
lighting than the yearly rate of casualties that this country 
suffered during the recent World War. According to statis- 
tics a few years ago an average of more than 100,000 men 
were continually incapacitated for work owing to accidents 
due to inadequate or improper lighting. This number of men 
could mine over 100 million tons of coal per year and only 
10,000 tons of it would be necessary to supply them with 
adequate and proper lighting. According to Simpson, in- VISIBILITY AND SAFETY 
285 
dustrial accidents cost this country at the present time about 
two billion dollars annually. Twenty per cent of this loss 
could be eliminated by providing good lighting. This sav- 
ing would pay the present artificial lighting bill for the 
entire country. Lighting will some day be a compulsory 
safeguard just as railings and other safety devices are today. 
Traffic 
The moving of traffic safely and rapidly is very largely 
a matter of good lighting. In walking or in driving we de- 
pend upon vision and it must be quick and certain. We 
have traffic in factories, in office buildings, on the streets 
and everywhere that there is human activity. Automobile 
accidents, according to Crum, cost this country about one 
billion dollars per year and it is fairly certain that nearly 
20 per cent are due to inadequate or improper street-light- 
ing. Crimes are found to be reduced on streets after they 
have been adequately lighted. 
But there is another aspect that has been generally over- 
looked and that is the bearing of adequate and proper light- 
ing upon the rate of movement of traffic. This does not 
mean the use of light-signals, for that is quite another mat- 
ter. For example, signs or other signals may direct a person 
downstairs but the speed at which the stairs will be de- 
scended (other things being equal) will depend in some 
manner upon the intensity of illumination, the position of 
the light-source, freedom from glare, etc. Certainly a 
person proceeds cautiously, even falteringly, in the dark. 
On increasing the amount of light, when is the point reached 
beyond which no increase in natural speed is noticeable? 
Certainly this point is not at the low intensities found on 
our streets and even in hallways, subways, etc. In order 
to get the greatest natural movement of traffic, the lighting 
must inspire a feeling of complete safety. The amount of 
light must be adequate. Shadow is necessary to reveal 286 
LIGHT AND WORK 
obstacles and steps if they are not revealed by painted 
contrast. Visibility should not be impaired by glaring light- 
sources in the field of view. The general stimulating effect 
of adequate light and bright surroundings should also be a 
factor in the movement of traffic. 
Visualizing these obvious requirements in direct com- 
parison with conditions as we find them in streets, in 
corridors, on stairways, in subways, and even in stores, 
certainly is convincing that here is a field for research and 
possible accomplishment. Street-lighting is almost every- 
where inadequate, but it involves problems and economic 
considerations which separate it from many of the other 
highways of traffic. There is little excuse for the inadequate 
lighting in corridors and on stairways where traffic is heavy; 
still these are usually less adequately illuminated than the 
other portions of the building. In our subways, on our ele- 
vated platforms and in our railway stations where the move- 
ment of traffic is a vital factor in the carrying capacity of 
cars (at least during the rush hours), adequate and proper 
lighting would be effective. There are also many stores in 
which the movement of traffic is a vital factor on sales. Here 
the lighting is usually much better than in the other places 
mentioned, but is it the best? Is it such that it inspires 
complete safety, provides the best visibility, and stimulates 
the greatest activity? These three goals must be achieved 
by lighting before the highest natural speed of the traffic 
is attained. Everywhere we look in our congested cities we 
see opportunities for light to hasten the footstep by inspiring 
confidence and stimulating activity. 
Traffic Signals 
The application of electric light in controlling and direct- 
ing traffic is growing rapidly and with this growth there is 
an increasing need for standardization. It is not surprising 
that self-luminous signals are finding favor for several rea- VISIBILITY AND SAFETY 
287 
sons. Such signals can operate at any time during night as 
well as day. They can be as powerful as desired, inasmuch 
as this is a function of the intensity of the light-source 
associated with them, admitting other things being equal. 
In the control of traffic, vision and its more common de- 
fects must be considered. Form and color are the chief 
fundamentals of a signal. These can be satisfactorily in- 
corporated in a self-luminous signal. Form alone is of 
very doubtful value because a signal should be recognizable 
at a reasonable distance. Quite an appreciable number of 
persons have vision so defective that they cannot accurately 
distinguish form at a distance. However, this difficulty is 
not found in the recognition of certain colors. There is a 
tradition to the effect that about four per cent of the people 
are color-blind but this statement is as misleading as it is 
indefinite. 
Only two signals are fundamentally necessary for STOP 
and GO respectively. Red and blue-green are quite suitable 
and it is exceedingly difficult to find a person who cannot 
distinguish these two colors as being different. Persons 
experienced in color-vision believe that perhaps no more 
than one in a hundred thousand persons would have diffi- 
culty in distinguishing these two colors as being different. 
However, it is likely that a thousand times more persons 
have eye-defects so serious as to make form unrecognizable 
at a reasonable distance. Furthermore, fog, rain, smoke and 
dust often obliterate form when under the same conditions 
a powerful self-luminous color-signal would still be seen as 
a spot of colored light. Even the selective transmission of 
fog, smoke, etc. cannot alter a blue-green light sufficiently 
at a reasonable distance to cause it to be confused with 
red. 
In practice a third color has been found desirable as an 
" interval " color; that is, as a signal to clear up matters 
before the next signal is flashed. Yellow is found satisfac- 288 
LIGHT AND WORK 
tory for this, for it offers no difficulties from the color-vision 
viewpoint. 
After utilizing color, the most powerful and practical 
aspect of light for traffic signals, it may be well to take 
advantage of form. With colored signals this may be largely 
a matter of position. For example, we may have a hori- 
zontal line of red lights and a vertical line of blue-green 
lights. Certainly many things can be done, but the crying 
need is to standardize upon one of the best and to adopt 
it generally. 
In signals, attention-value is important and the two colors 
red and blue-green seem to meet the requirements of ap- 
propriateness and distinctiveness. Intensity, of course, is 
another factor upon which attention-value depends, but the 
mistake is often made of depending upon photometric meas- 
urements or candle-power for determining this phase of 
attention-value. Here we encounter something akin to glare 
and certainly glare finally must be measured by the eye. 
Let us take as an example the so-called " stop-light " com- 
monly used on automobiles. Let us assume two lenses 
equal in diameter and of the same candle-power toward the 
eye. If one of these is a ground glass and the other is 
covered with indentations of appreciable size, the latter will 
be found to be of greater attention-value, even when all 
other factors such as color, size, and candle-power are equal 
respectively for the two signals. Of course, this is not true 
when the two signals are so far away as to appear as points 
of light. The explanation lies in an unexplored realm of 
physiological optics, but an analogy throws some light upon 
the matter. For example, if we wish to awaken someone by 
throwing something against the window pane, a handful of 
sand is not as effective as an equal weight of coarser peb- 
bles. In fact, one large stone equivalent in weight to the 
handful of sand would likely break the glass while the hand- 
ful of sand would be harmless in this respect. So it is with VISIBILITY AND SAFETY 
289 
light. A certain intensity of light from a ground glass is 
not as glaring or as powerful in gaining attention as the 
same intensity from a glass of equal area but consisting of 
larger bundles of light-rays such as emitted by a glass con- 
sisting of proper indentations of appreciable size. 
These are important factors in traffic signals which must 
be understood in all their variety of influences on vision if 
the best choice of signals is to be made. It is to be hoped 
that our greatly increasing traffic will result in a special study 
and proper application of light. When this is done self- 
luminous colored signals having lenses or beams of proper 
optical design will be found best and this should be 
universally standardized. REFERENCES 
The following list includes only those references from which material 
was drawn or in which particularly pertinent data are to be found. 
It is not to be considered an adequate bibliography of light, color, 
and vision. 
1. Color and Its Applications, M. Luckiesh, 1921. 
2. Ultraviolet Radiation, M. Luckiesh, 1922. 
3. Sky Brightness and Daylight Illumination, H. H. Kimball, 
Trans. I.E.S. 16, 1921, 255; 18, 1923, 435. 
4. The Sun, C. G. Abbot, 1911; Ann. Astrophys. Obs. vols. II 
and III. 
5. The Physics of the Air, W. J. Humphreys, 1920. 
6. Light and Shade and Their Applications, M. Luckiesh, 1916. 
7. Daylight Illumination, H. H. Kimball, Trans. I.E.S. 11, 
1916, 397. 
8. Visibility of Airplanes, M. Luckiesh, Jour. Frank. Inst. 187, 
1919, 289; Aerial Photometry, Astrophys. Jour. 49, 1919, 108. 
9. Light and Color in Advertising and Merchandising, M. 
Luckiesh, 1923. 
10. Pathological Effects of Radiant Energy on the Eye, Verhoeff 
and Bell, Proc. Amer. Acad. Arts and Sci. 51, 1916, 635. 
11. The Cost of Daylight, Luckiesh and Holladay, Trans. I.E.S. 
18, 1923, 119. 
12. Measurement of Transmission-Factor, Luckiesh and Mellor, 
Jour. Frank. Inst. 186, 1918, 529. 
13. Some Principles Governing the Proper Utilization of Light in 
Roof Fenestration, W. S. Brown, Trans. I.E.S. 19, 1924. 
14. Daylighting from Windows, Higbie and Younglove, Trans. 
I.E.S. 19, 1924. 
15. Determination of Daylight Intensity at a Window Opening, 
H. H. Kimball, Trans. I.E.S. 19, 1924. 
16. Planning Buildings for Daylight, Swan and Tuttle, Architec- 
tural Forum, November, 1918; Sunlight Engineering Architectural 
Forum, June, 1918. 
17. Artificial Light - Its Influence on Civilization, M. Luckiesh 
1920. 
290 REFERENCES 
291 
18. Artificial Daylight, M. Luckiesh, Trans. I.E.S. 9, 1914, 839; 
Elec. World, 64, 1914, 570; 66, 1915, 71. 
19. Influence of Colored Surroundings on the Color of Useful Light, 
M. Luckiesh, Trans. I.E.S. 8, 1913, 61. 
20. Chromatic Aberration in the Eye, Dr. Louis Bell, Elec. World, 
57, IQ", 1163. 
21. Monochromatic Light and Visual Acuity, M. Luckiesh, Elec. 
World, 58, 1911, 450 and 1252. 
22. Data Pertaining to Desired Illumination Intensities, Luckiesh, 
Taylor and Sinden, Jour. Frank. Inst. 192, 1921, 757. 
23. The Spectral Distribution of Energy Required to Evoke the 
Gray Sensation, I. G. Priest, #417 Sci. Pap. of Bur. of Stds. 
24. A Study of Natural and Artificial Light Distribution in In- 
teriors, M. Luckiesh, Trans. I.E.S. 7, 1912, 388. 
25. Effective Lighting of Factories as Judged by Daylight Stand- 
ards, Ward Harrison, Trans, I.E.S. 12, 1917, 409. 
26. Influence of Daylight Illumination intensities on Electric Cur- 
rent Used for Lighting Purposes, A. Smirnoff, Trans, I.E.S. 19, 1924. 
27. Coefficients of Utilization, E. A. Anderson, Trans. I.E.S. 15, 
1920, 97. 
28. The Lighting Art, M. Luckiesh, 1917. 
29. Introduction to the Study of Color-Vision, J. Herbert Parsons, 
I9I5- 
30. Reaction of the Eye to Light, P. Reeves, Trans. Opt. Soc. 22, 
1920-1921, 1. 
31. Retinal Sensibilities, P. G. Nutting, Trans. I.E.S. n, 1916, 1. 
32. Effect of Size of Stimulus on the Contrast Sensibility of the 
Retina, P. Reeves, Jour. Opt. Soc. Amer. 1, 1917, 148. 
33. Influence of Pupillary Aperture on Visual Acuity, P. W. Cobb, 
Amer. Jour. Psych. 36, 1915, 335. 
34. Response of the Average Pupil to Various Intensities of Light, 
P. Reeves, Jour. Opt. Soc. Amer. 4, 1920, 35. 
35. Dark Adaptation of the Human Eye, Selig Hecht, Jour. Gen. 
Psych. 2, 1920, 499. 
36. Growth and Decay of Color-sensations in Flicker Photometry, 
M. Luckiesh, Physical Review 4, 1914, 1. 
37. Jour, de Physiol, et de Path. Gen. No. 4, July, 1902. 
38. Incandescent Lamp Temperatures as Related to Modern 
Lighting Practice, C. L. Dows and W. C. Brown, Trans. I.E.S. 16, 
1921, 284. 292 
LIGHT AND WORK 
39. Depreciation of Lighting Equipment Due to Dirt and Dust, 
Anderson and Ketch, Trans. I.E.S. 19, 1924, 55. 
40. Maintenance of the Lighting System, A. L. Powell, Bull. Edison 
Lamp Works, April, 1920. 
41. Illumination Design Data, E. A. Anderson, Bull. 41 National 
Lamp Works of G. E. Co. Jan. 15, 1921. 
42. Visual Illusions, M. Luckiesh, 1922. 
43. The Relation of Illumination to Production, D. P. Hess and 
Ward Harrison, Trans. I.E.S., 18, 1923, 787. 
44. Effects of Brightness and Contrast in Vision, P. G. Nutting, 
Trans. I.E.S. 11, 1916, 939. 
45. Relation between Field Brightness and Speed of Retinal Im- 
pression, P. W. Cobb, Jour. Expt'l. Psych. 4, 1923, 138. Some 
experiments on Speed of Vision, Trans. I.E.S. 19, 1924, 150. 
46. Glare Measurements, Ward Harrison, Trans. I.E.S. 15, 
1920, 34. 
47. Report of Sub-Committee on Glare, Louis Bell, L. T. Troland, 
F. H. Verhoeff, Trans. I. E. S. 17, 1922, 743. 
48. High Cost of Poor Lighting, R. E. Simpson, Trans. I.E.S. 
15, 1920, 576. 
49. Trans. I.E.S., 17, 1922, 76. 
The Code of Lighting Factories, Mills and other Work-Places 
and also The Code of Luminaire Design prepared and issued by the 
Illuminating Engineering Society, New York are valuable aids in 
lighting-practice. INDEX 
Absorption of light, 218 
Accidents, 201, 284 
Adaptation, 132, 135, 145, 163 
After-image, 168, 169 
Anderson, E. A., 228, 234 
Arc lamps, 76 
Artificial daylight, 100 
Artificial illuminants, 96 
Artificial light, 71 
cost of, 81 
Atmosphere, 24 
Average of reflection-factor, 257 
Ciliary body, 131 
Cleaning glass, 216 
Closing pupil, 162 
Clouds, 27, 32 
Sunlight, 282 
Cobb, P. W., 193, 157, 184, 255 
Color, 86, 97 
discrimination, 155 
saturation of, 154 
sensation, 152 
work, 98 
of illuminants, 96 
for ceiling, 235 
for walls, 235 
Commery, E. W., 262 
Cones, 133 
Confusion pattern, 193, 256 
Conjunctiva, 131 
Contrast, 254 
excessive, 280 
simultaneous, 155 
Cost of artificial light, 81 
daylight, 62 
depreciation, 222 
food, 83 
installation, 263 
light, 121, 262, 266 
operating, 263 
per outlet, 263 
Crime, 285 
Crum, 285 
Background, 174 
Bell, Louis, 115 
Binocular vision, 172 
Biological effects, 14 
Blanchard, 250 
Blind-spot, 131, 152 
Blue sky, 90 
Bowl-enamelled, 282 
Brightness, 240 
-difference, 139 
excessive, 280 
of filament, 169 
of light-sources, 282 
of sky, 32, 34 
sensation, 138 
-units, 137, 242 
Broca & Sulzer, 167 
Buildings, windowless, 64 
Candle-power, 21 
-flame, 90 
Carbon arc, 282 
Carbon-dioxide tube, 93 
Carbon filament, 90 
Cast shadow, 173 
Ceiling, 226 
color, 235 
effective, 229 
Choice of intensity, 117 
Choroid, 131 
Chromatic aberration, 115 
Daylight, artificial, 100 
cost of, 62 
equipment, 216 
indoors, 47, 126 
intensity, 22 
lamp, 90, 108 
openings, 48 
outdoors, 20, 127 
Dazzle, glare, 282 
Decay of sensation, 166 
Demonstrating speed of vision, 196 
Depreciation, 265 
293 INDEX 
294 
cost of, 222 
artificial light, 219 
Desirable intensity, 246, 260 
Difference of brightness, 139 
Diffuse reflection, 231 
Diffusing glass, 282 
Discipline, 201 
Discomfort, 209 
Discrimination, color of, 155 
speed of, 178 
Distinguishable hues, 154 
shades, 154 
tints, 154 
Dull paint, 230 
Duration of after-image, 169 
sunshine, 23 
Durgin, W. A., 202 
Flames, 106 
Flicker, 170 
frequency, 169 
Foot-candle, 240 
Form of curves, 271 
Fovea, 131, 133, 152 
Fundamental lighting conditions, 15 
Fundamentals of vision, 130 
Gas light-sources, 71 
Glare, 164, 238, 280 
dazzle, 282 
veiling, 281 
Glasses, 65 
diffusing, 282 
Glass prisms, 69 
Globe diameter, 283 
Glossy paint, 231 
Growth of sensation, 166 
Earth areas, reflection-factors of, 
38 . 
Economic intensity, 261 
Effective ceiling, 229 
intensity, 240 
sky area, 54 
Effect of walls, 229 
Electric filament lamps, 78 
light-source, 75 
Energy, spectral distribution of, 89 
Environment, 3 
Esthetic self, 106 
Evolution of eye, 9 
Excessive brightness, 280 
contrast, 280 
Exposure, 281 
External factors of vision, 171 
Eye, 130 
comfort, 279 
fatigue, 107, 283 
lens, 157 
strain, 209, 283 
Harrison, Ward, 124, 204, 220, 270 
Hecht, 164 
Hess, D. P., 204, 270 
Highlight, 174 
Holladay, 272 
Hothouse effect, 58 
Hue, 153 
distinguishable, 154 
Hygienic conditions, 201 
Hygienic value of radiation, 60 
Illuminants, 86 
artificial, 96 
color of, 96 
natural, 94, 96 
Illumination intensity, 51, 207 
and vision, 247 
Increased production, 202, 265 
Indoor daylight, 126, 47 
Infra-red radiation, n, 59, 238 
Installation cost, 263 
Intensity, choice of, 117 
desirable, 246, 260 
effective, 240 
of illumination, 51, 207 
of light, 123 
Invisible radiation, reflection of, 238 
Iris, 131 
Irradiation, 156 
Factors of glass transmission, 69 
of safety, 248, 255 
Fading of materials, 61 
Ferree & Rand, 179 
Fine detail, 178 
Filament brightness, 169 
electric lamp, 78 
Fire, 71 
Fixed charges, 264 
Jones, L. A., 154 INDEX 
295 
Kdnig, 139, 165 
Opening pupil, 162 
Openings, daylight, 48 
Operating cost, 263 
Optic nerve, 131 
Outdoor daylight, 20, 127 
Overshooting of sensation, 167 
Lamp, daylight, 108 
renewals, 263 
Landscapes, 37 
Lens, 131 
of eye, 157 
Light, absorption of, 218 
artificial, 71 
cost of, 206 
intensity of, 123 
and payroll, 270 
psychology of, 105 
quality of, 56, 86 
quantity of, 119, 281 
Lighting principles, 41 
and production, 199 
results of, 17, 200 
value of, 212 
Lighting value of paint, 225 
Light-sources, brightness of, 282 
electric, 75 
gas, 71 
primitive, 73 
temperature of, 90 
Living things, 5 
Logarithmic curves, 271 
Low-contrast, 254 
Lumens per watt, 92 
Luminous efficiencies, 79, 92 
Paint, 225 
Paint, lighting value of, 225 
white, 233 
Painted machinery, 237 
Payroll, 267 
and lighting, 270 
Pfund, 238 
Pigments, 235 
Polished surfaces, 281 
Population of earth, 120 
Powell, A. L., 223 
Primitive light sources, 73 
man, 6 
Prism glass, 69 
Production, 202 
increase of, 202-265 
and lighting, 199 
tests, 202 
Properties of radiation, 12 
Psychology of light, 105 
Pupil, 156 
size of, 161 
Purkinje, 154 
effect, 165 
Machinery, painted, 237 
Magee, John, 203 
Magnesium oxide, 238 
Maintenance, 215, 263 
Manufacturing statistics, 201, 210 
Mercury arc, 77, 90, 117 
Millilambert, 242 
Mirror, 231 
Monochromatic light, 115 
Moonlight, 35 
Moore tube, 93 
Quality of light, 56, 104, 186 
Quantity of light, 119, 281 
Radiation, biological effects of, 14 
hygenic value of, 60 
infra-red, 238 
properties of, 12 
solar, 8 
ultraviolet, 11, 238 
visible, 10 
Reading, speed of, 187 
Reeves, 160, 250 
Reflecting characteristics, 230 
Reflection, diffuse, 231 
regular, 231 
factor, 140, 227, 232, 240 
factor average, 257 
factors of earth areas, 38 
Natural illuminants, 94, 96 
quality, 105 
North light, 99 
Nutting, P. G., 154, 248 
Oil lamps, 73 296 
INDEX 
of invisible radiation, 238 
from surroundings, 55 
Refraction, 69 
Refractive index, 69 
Regular reflection, 231 
Resolving power, 157, 135 
Results of improved lighting, 200 
of lighting, 17 
Retina, 131, 152 
Revealing power, 118 
Rods, 133 
Surroundings, reflection from, 55 
Suspensory ligament, 131 
Taylor, A. H., 234 
Temperature of light sources, 90 
Test-objects, 136 
object international, 180 
Tint, 153 
distinguishable, 154 
Traffic, 285 
signals, 286 
Transmission-factors of glasses, 69 
Tungsten filaments, 282 
lamp, 90, 117 
Safety, 209, 279, 284 
Saturation of color, 154 
Sclera, 131 
Sensation brightness of, 138 
color of, 152 
decay of, 166 
growth of, 166 
Shade, 153 
distinguishable, 154 
Shadow, 141, 143, 150, 173 
Signals, traffic, 286 
Simpson, R. E., 284 
Simultaneous contrast, 155 
Size of pupil, 161 
Sky brightness, 32, 34 
light, 22, 28, 31 
lights, 49 
Smirnoff, 126 
Snow, 56 
Solar radiation, 8 
Spectral character, 87 
distribution of energy, 89 
luminosity curves, 88 
regions, 13 
Spectrum visible, 10 
Specular reflection, 232 
Speed of reading, 187 
discrimination, 178 
vision, 176 
Spoilage, 201 
Starlight, 35 
Stickney, G. H., 197 
Sun,120 
light, 20, 22, 31, 88, 90 
Sunlight cloud, 282 
Sunshine, duration of, 23 
Surfaces, polished, 281 
Ultraviolet radiation, n, 238 
Units, brightness of, 137, 242 
Unnatural light, 113 
Utilization factor, 226 
Value of lighting, 212 
reflection, 227 
Variation of daylight, 28 
Veiling glare, 281 
Visibility, 115, 279, 284 
Visible radiation, 10 
spectrum, 10 
Vision, binocular, 172 
demonstrations of speed, 196 
external factors of, 171 
fundamentals of, 130 
and illumination, 247 
speed of, 176 
Visual acuity, 115, 134, 147, i49> T79 
angle, 132, 146 
data, 251 
functions, 244 
Wall color, 235 
Walls, 226 
effect of, 229 
Water vapor, 59 
White paint, 233 
Windowless buildings, 64 
Windows, 48, 54 
Working self, 106 
World War, 284 
Yellow light, 107 
Zinc oxide, 238 OTHER BOOKS 
BY 
M. LUCKIESH COLOR AND ITS APPLICATIONS 
Second Edition, Revised and Enlarged. 
6x9,150 illustrations, 4 color plates, 431 pages.$4.50 
The object of this treatise is not only to discuss the 
many applications of color, but to establish a sound scientific 
basis for these applications. The book is authoritative, 
well illustrated, and contains many references and a wealth 
of new material. It was written by an investigator in the 
general field of color and is therefore not narrowly limited 
in scope. It fills a distinct gap that has existed on the 
book shelves. 
LIGHT AND SHADE AND THEIR APPLICA- 
TIONS 
6x9, 135 illustrations, 277 pages $3.00 
The book is a condensed record of several years' research 
by the author in the science of light and shade. It is the 
first published work which deals with the science of light and 
shade in a complete and analytical manner. The book is of 
extremely wide interest because it deals with the appearances 
of objects and hence with vision and with lighting. It is 
well illustrated and represents the first elaborate attempt to 
formulate the science of light and shade and to correlate it 
with various arts. 
VISUAL ILLUSIONS, THEIR CAUSES, 
CHARACTERISTICS AND APPLICATIONS 
6x9, 100 illustrations, 258 pages $3.00 
There are numberless visual illusions, all of them inter- 
esting but many can be put to useful service in daily life. 
In this book will be found a condensed treatment of the 
practical aspects of visual illusions. The book emphasizes 
experimental facts and introduces theoretical considerations 
occasionally but chiefly for illustrating explanations which 
otherwise would be too complex. ULTRAVIOLET RADIATION; ITS PROPERTIES, 
PRODUCTION, MEASUREMENT, AND APPLI- 
CATIONS 
6x9, illustrated, 249 pages $3.50 
This book presents authentic data useful to those who 
are interested in ultraviolet radiation. Theory has been 
subordinated to experimental facts pertaining to properties, 
production and applications. The material and references 
presented should be of practical value to chemists, 
physicists, engineers, biologists, ophthalmologists, physicians 
and others. 
LIGHT AND COLOR IN ADVERTISING AND 
MERCHANDISING 
5| x 8j, 30 multi-colored and 8 one-color plates 
280 pages $3.00 
This book presents a popular practical treatment of the 
selling power of light and color, and shows how these 
wonderful tools can be used to move merchandise through 
magazine advertising, posters, billboards, show-windows, 
distinctive shops, stores, electric signs and other channels. 
Years of research and analysis by the author has yielded 
a mass of data which for the first time has been coordinated 
for the use of the advertising specialist, the commercial 
artist, the color printer, the merchandiser, the architect, the 
interior decorator, the lighting specialist, the display man, 
the window trimmer, the manufacturer, and many others. 
Light and color are silent, inexpensive, powerful salesmen 
that should not be ignored by any one engaged in business. 
PORTABLE LAMPS; THEIR DESIGN AND USE 
5| x 8|, illustrated, 144 pages $2.00 
The use of the portable lamp is rapidly growing. Many 
lighting problems in the home and elsewhere can be suc- 
cessfully solved by properly designed portables. They 
satisfy artistic requirements and perform various utili- 
tarian functions. There are many scientific and artistic 
principles involved in portable lamps with which the user 
is not generally acquainted. This book is a valuable aid in 
designing, in purchasing and in using portable lamps. LIGHTING THE HOME 
5 x 7|, illustrated, 289 pages $2.00 
This is a pioneer book. It ranks with books on interior 
decoration and furniture as a help toward transforming a 
house into a home. It is practical in that it offers advice 
on all sorts of lighting problems and it is fascinating reading 
as well. 
ARTIFICIAL LIGHT, ITS INFLUENCE ON 
CIVILIZATION 
6x9, illustrated, 366 pages $3.00 
This story of the achievements of artificial light is written 
especially for the man in the street who is not interested 
in technical scientific terms and formulae, but who looks 
with admiration upon the huge signs which flash and 
sparkle above the crowds on the Great White Way, who 
marvels at the colors and brilliance of a spectacular theatri- 
cal production and desires to know how it is accomplished, 
and who takes a natural delight in hearing about scientific 
discoveries when they are explained in the simple, vivid 
language he understands best. 
THE LIGHTING ART, ITS PRACTICE AND 
POSSIBILITIES 
6x9, illustrated, 229 pages $2.50 
This book discusses lighting as engineering plus art, and 
treats the subject as a branch of interior and exterior 
decoration. The technical aspect of the subject is not 
neglected, but the main emphasis is upon the " why " and 
not merely the " how " of lighting. 
THE LANGUAGE OF COLOR 
6x9, illustrated, 282 pages $2.00 
A practical volume on color, the various fields in which 
it is used and its importance in portraying the ideas that 
make for progress. A book of special interest to all those 
who deal in color schemes and values. 
THE BOOK OF THE SKY 
6x9, illustrated, 236 pages $3.50 
"The beauties, wonders, awesome spectacles, inspiring 
panoramas, and extensive ranges of vision which await 
the aerial traveler, make of cloudland a veritable fairyland 
if he will open his consciousness to them. Aircraft have 
brought this new world of experiences within easy reach 
of mankind and it is one of the aims of this volume to 
awaken those who fly, or would fly, to the variety of 
interest which air travel affords." The 
Literature of Engineering 
and the Industries 
On our shelves is the most com- 
plete stock of engineering, scientific, 
technical and industrial books in the 
United States. A large number of 
these we publish and for an ever in- 
creasing number we are the sole 
agents. 
<J All inquiries are cheerfully 
and carefully answered and com- 
plete catalogs as well as special 
lists are sent free on request. 
D. Van Nostrand Company 
PUBLISHERS AND BOOKSELLERS 
8 WARREN STREET NEW YORK