[From the Am. Journal of Science and Arts, Yol. X, Supplement No., 1875. | 
THE SOLAR .ATMOSPHERE. 
AN INTRODUCTION TO AN ACCOUNT OF RESEARCHES MADE AT 
THE ALLEGHENY OBSER/ATORY. 
By 8. P. LANGLEY, 
DIRECTOR OF THE OBSERVATORY. 
Paper read at the Detroit meeting of the American Association for the Ad- 
vancement of Science. 
It has long been observed that the snn is not everywhere 
equally bright, but that the edge is darker than the center, and 
since a self-luminous sphere should appear of sensibly uniform 
brilliancy, so as to present to the eye the appearance of a flat 
disc, this diminution of light must be due to some medium 
external to the photosphere. Accordingly, even before the in- 
vention of the spectroscope, it was admitted that an absorbing 
atmosphere surrounded the sun and that the effect would neces- 
sarily be to diminish the amount of radiation. La Place, in 
the tenth book of the Mecanique Celeste, has attempted to com- 
pute the total effect of this absorption, from data furnished by 
the observations of Bouguer, and considers that were the solar 
atmosphere removed, the heat and light would be twelve times 
as great as at present. Though this value is erroneous, being 
deduced from imperfect data by processes which rest on a false 
hypothesis, it may yet serve to call attention to the importance 
of a somewhat neglected field of research, in which this obser- 
vatory has been in part occupied during the past three years. 
Several estimates of the absorptive power of this atmosphere 
have been made, differing widely from each other. According 
to that of La Place just cited the absorption is about {-£ of the 
sun’s emission, according to Liais less than ; according to 
Secchi ’88. These discrepant results may be due in part to 
different hypotheses used in computation, but they are in every 490 
S. P. Langley—The Solar Atmosphere. 
case founded on an experimental comparison of the light or 
heat of the sun observed at the center, with that observed near 
the edge. 
This direct comparison of the two lights involves, of course, 
no hypothesis, but is simply a photometric measurement, and 
apparently an easy one. Let us examine the result of this com- 
parison at different hands. Arago* using the Rochon prism 
and analyzer announced as the conclusion of a prolonged re- 
search, that the light at the center of the solar disc must be di- 
minished by y'r part, or 2’4 per cent to equal that of the edge. 
Liais,f announces from comparisons carried on by the extinction 
of one light by a stronger, that the central light should be di- 
minished by about 10 per cent to equal that of the edge. His 
result is then four times as great as Arago’s. Secchi, with a wheel 
photometer, finds for a nearly corresponding point *78 per cent 
as the amount by which the central light should be diminished. 
This result is then seven times that of Liais, and thirty times 
that of Arago. 
Secchi has made observations on radiant heat to be found in 
Le Soleil\ the most important results of which are that the heat 
diminishes from the center to the edge; that for a given point 
there is a satisfactory agreement between the absorption of light 
and heat; and that the equatorial regions are hotter than the 
polar. The present writer has made somewhat extended re- 
searches in the same direction which are chiefly unpublished. 
Yogel has made interesting researches in the relative actinic ab- 
sorption. 
None of those who have attempted to compute the amount 
of the solar absorption appear to have drawn conclusions from 
their results as to its effect on terrestrial temperatures, and yet 
if the absorption be anything like what has been found by La 
Place, and by Secchi, using La Place’s formulae, the subject de- 
serves attention. 
It will be shown in a forthcoming memoir that the absorption 
is much less than these values, but that definite limits can be 
assigned, within which it must lie, and that at any rate it is 
of such importance that the conditions of animal life upon 
tbis plenet largely depend upon it. It will further appear 
probable that with a slight change in the depth and absorptive 
power of this atmosphere, fluctuations in terrestrial temperature 
will ensue, very great in comparison with any actually observed 
within historic periods, and it will be shown that this atmos- 
phere is not in a strictly stable condition. Though the subject 
* Oeuvres Completes, Tom. i, p. 235. 
t Memoires de l’Academie de Cherbourg. 
X Le Soleil, 2me edition, Paris. 1875. S. P. Langley—The Solar Atmosphere. 
491 
then is a nearly neglected one, and the methods of investiga- 
tion of less apparent interest than those of the spectroscopist, 
the results are, if verified, of a peculiar interest, from their 
bearing upon the possible changes in the mean temperature 
and climatic conditions of our own planet. 
The portion of this atmosphere chiefly concerned in absorp- 
tion I have been led to believe from several considerations is 
extremely thin, and I am inclined to think it is nearly identi- 
cal with the “ reversing layer” at the base of the chromosphere 
observed by Secchi and Young, though the other chromospheric 
strata doubtless, have some share in the obscuration made, 
and, in a still less degree, the other solar envelopes. 
The methods used for heat measurements at Allegheny have 
been partially described,* and need not be enlarged upon here. 
Those for light are believed to be novel in their present appli- 
cation, and may be given briefly. 
We have seen that a photometric comparison of the center 
and edge of the sun is for some reason attended with risk of 
error which we should not anticipate, but which must be great 
from the enormous discrepancies existing among practiced ob- 
servers ; for, taking Arago’s value as the standard, Liais’ is four 
hundred per cent, and Secchi’s three thousand per cent greater. 
Now skilled observers, in comparing their estimates of the ex- 
cess of the light of a gas flame over that of a candle, for in- 
stance, by the familiar methods of ordinary photometry, need 
not be expected, as common experience shows, to commit er- 
rors which are in any way comparable to these. 
If such familiar means as the Rumford and Bunsen photo- 
meters could be used for the direct comparison of the solar in- 
tensities there seems no reason why the results should not be of 
the exactness obtainable in the physical laboratory, or at least 
of the same order of accuracy. If results can be obtained by 
such means, they may at least be hoped to be free from gross 
error. 
At first sight it appears that these methods are not applica- 
ble to the direct comparison, for the ordinary use of either pho- 
tometer supposes that the relative distances of the lights from 
the screen can be altered, while we cannot change the relative 
distances of the lights to be here compared, which are two por- 
tions of the sun; and, if we attempt to enfeeble the stronger 
lights by shades, or by diffusing it through a lens till it equals 
that of the other seen direct, we lose the peculiar advantages 
of the methods which assume that the lights are viewed under 
the same conditions and measured by the relative distances 
from the screen. 
I arranged, in June, 1874, an apparatus which appears to be 
free from these objections, and which is here described in 
* Oomptes Rendus, May 22, 1875. 492 
S. P. Langley—rIhe Solar Atmosphere. 
principle, not in detail. Let us suppose that to the equatorial 
is attached a screen upon which is described a large circle 
whose radius is divided into 100 equal parts, and whose center 
is always in the prolongation of the optical axis. This screen 
receives an enlarged solar image from an amplifying lens. To 
fix our ideas we may suppose that the circle and image are 
each always 24 inches in diameter, and that a direct comparison 
is to be made of the light of the center with that of a point £ 
of the way from the center to the western edge of the sun. 
Two small* and equal circular apertures are made in the 
screen at points equidistant from the center or 75 divisions of 
the scale apart. The telescope is directed to neither point un- 
der examination, but to a point of the solar disc £ of the way 
to the western edge from the center, and which is therefore 
midway between the points to be measured. The image of the 
center now falls on one of the small apertures in the screen ; 
that of the point under examination on the other, and through 
the apertures pass two cones of rays each slightly divergent, 
formed under precisely the same conditions, since they come 
from the same lens and are equidistant from the optical axis. 
Two prisms of total reflection (cut from the same piece of 
glass) are placed one behind each aperture, and these deflect 
the rays toward each other, so that the reflected portion of the 
cones have a common axis behind and parallel to the screen, 
within a chamber which is lined with black and shielded from 
light, so as to form in fact a camera obscura. There is a small 
Bunsen disc behind the screen and in the camera, whose center 
remains in the axis common to the reflected cones as it (the 
disc) slides back and forth on a graduated scale which meas- 
ures the distance from the source of illumination. Evidently 
one light is diffused and the other concentrated by the disc’s 
advance and recession upon the cones, sensibly as though these 
were formed by lights at their virtual apices beneath the screen, 
and by lights which were of the color and intensity of the 
portions of the sun under examination. In practice it is found 
convenient to introduce a lens of short focus between each ap- 
erture and its reflector, so that the cones, while still exactly 
similar, may be less acute. By the substitution for the Bunsen 
disc, of a small box, sliding also on the graduated scale, and 
containing a rod whose shadows are cast, by means of a second 
pair of reflecting prisms, on a prepared surface, sliding with 
the box, we convert the instrument into a Rumford Photome- 
ter, and in this form also it has been much used here. This 
application incidentally permits under very favorable circum- 
stances a comparison of the color of the light from different 
parts of the sun, which is by no means uniform. When light 
* The diameter of the aperture is on the above scale about 0'-10". IS. P. Langley—The Solar Atmosphere. 
493 
from near the edge is juxtaposed with that from the center, the 
shadow illuminated by the former is chocolate-red,* while that 
from the latter is a peculiar bluish tint, nearly like that given 
by the shade-glass called by English opticians “London 
Smoke.” (I use these comparisons not being able readily to as- 
similate these colors to any of the pure spectrum.) It is worth 
while to remark that this observation confirms others obtained 
from different means by the writer who has elsewhere an- 
nounced the fact of a selective absorption in the solar atmos- 
phere of such a nature that were the envelope removed the gain 
in light would be greater than the gain in heat. The light we 
receive from the sun and universally call “ white” is thus seen 
to be, upon the whole, less refrangible than that which the sun 
would emit if deprived of its atmosphere; in which case it is 
evident that the color of our luminary as it grew brighter 
would tend toward blue. If then the depth of its atmosphere 
were sufficiently increased, our sun, in growing darker, would 
also appear more red. To more than suggest the possible 
influence of such a cause on the colored stars, among which the 
bluish are distinguished by the absence of absorption spectra, 
and the desirability of ascertaining whether change in color, if 
any such are verifiable, are or are not accompanied by slight 
changes in apparent magnitude, would lead us away from the 
present subject. 
So clear an exhibition of color in the actual comparison is a 
sign of the delicacy of the Euinford method, but it introduces a 
disturbance in our estimation of the intensities of colored 
lights. An arrangement of the Masson photometer (in which 
black and white sectors on a rapidly revolving disc illuminated 
by the colored lights are viewed by the intermittent electric 
discharge) has been prepared to obviate this; but the complete 
examination is intended to include that of juxtaposed spectra 
formed from the lights under examination, selected portions of 
which of different wave lengths will be compared, as regards 
intensity, by processes which are a development of those men- 
tioned above. In all such comparisons the apparatus is rever- 
sible, so as to eliminate any inequalities in the material, posi- 
tion, or condition of the rays under examination due to the in- 
strument ; and in practice corrections for every form of instru- 
mental error are applied which are not here mentioned. 
The methods above indicated are, with some slight modifi- 
cations, evidently adapted to the comparison of the light of the 
sun-spots, and the light adjacent to the limb, with that of the 
center ; to studying the rate of diminution of light without the 
disc, and to like investigations ; and they are now being so used. 
* This color appears to have passed unnoticed by all observers but Secchi, who 
remarks that the edge of the disc is smoky-red. 494 
S. P. Langley—The Solar Atmosphere. 
Without entering into detail, I may observe that the measure- 
ments of the light of spots, though as yet incomplete, warrant 
me in stating that the absorption of the more refrangible rays, 
though great in reference to the less refrangible, is not so great 
as would appear from published measurements. It has long 
been supposed that the umbrae of spots were not absolutely 
dark, while it has been admitted, from Dawes’s observation, that 
there is within the umbra a sensibly black “ and 
from the assumed blackness of this “nucleus,” several astron- 
omers have been led to suppose that the “ ” are not 
openings into a dark gaseous interior. 
The measurement of umbrae, and of so-called “ here 
shows not only that neither are absolutely dark, but that the ab- 
solute light of either is enormous, that of the average “ 
so-called, being, as I find, at least five thousand times that of the 
full moon. 
La Place, in assuming that the radiation in any direction is 
constant, and is proportional to the radiating area (or that the 
radiation is infinite at the edge of the disc), concludes that the 
total absorption is represented by the integral 
2 *f£ cosO ginOdO 
o 
where 6 is the heliocentric angle between the earth and the 
point of the solar surface under examination ; and this assump- 
tion has been adopted by Father Secchi, who, in his latest 
edition of Le Soleil, gives results derived from the use of this 
integral. I shall, however, assume, in accordance with what 
seem to be the teachings of modern physics, that a globe as 
large as the sun whose photosphere, though composed perhaps 
of very light vaporous material, is yetopake at a limited depth, 
(as observations on superposition of cloud-strata have shown) 
that in such a globe radiation would be proportional to the co- 
sine of the angle between the normal to the surface at any 
point and a line drawn from that point to the observer’s posi- 
tion ; or, that the sun, deprived of its atmosphere, would appear 
as a flat disc ; whence it will follow that, the total radiation of 
the sun without the atmosphere being unity, that which es- 
capes through the atmosphere will be, if we adopt La Place’s 
. 1 ~ 
formula with this correction 
2zr / £ cos6 cosfeinOdd. 
v o 
These expressions, if La Place’s assumptions are otherwise 
correct, should give, for any comparison of the heat at a given 
point of the disc to that of the center, a certain value of the 
absorption which should be constant for any value of 6. In 
fact I find, however, that this is not the case. The diserepan- S. P. Langley—The Solar Atmosphere. 
495 
ciesin the summation I obtain for different values are not how- 
ever casual and irregular, as they would be if due to errors of 
observation, but are systematic. This formula even as here 
modified, though analytically speaking correct, yet appears to 
rest on assumptions which are in disaccordance with facts ascer- 
tained since La Place wrote. These facts, as they seem to me, 
demanding a changed formula I expect to point out in a defi- 
nite form in a subsequent memoir. It seems fitter then to 
defer an exact statement of the limits of the amount by which 
solar radiation is absorbed by the atmosphere, until it is accom- 
panied by demonstration. I will only here repeat that we may 
feel certain that the estimates cited from La Place and Secchi, 
and which make the sun’s atmosphere absorb from f to ji of 
the radiation we should receive in its absence, are in excess of 
the truth. For the sake of indicating approximately the real 
value, I may also state that from my computation of the so- 
called “luminous-heat” rays, not greatly less or more than one- 
half the whole are absorbed by, turned back by, or converted 
into work in, the sun’s atmosphere. The total thermal absorp- 
tion is somewhat less than that of the luminous heat. If, how- 
ever, we admit this value provisionally, certain results seem to 
follow which deserve mention. 
The mean surface temperature of our globe is separated from 
that of absolute zero by about 500° of the Fahrenheit scale. 
The internal heat supplied to the surface may be neglected. 
The temperature of interplanetary space is, according to Fou- 
rier — 60°C., according to Pouillet —142, according to Liais —97. 
Liais, in a memoir whence we cite these estimates, admits, as a 
consequence of his own, that the obscure heat received at the 
upper 'Hrrrb-of our atmosphere from space (that is, from the 
chiefly non-luminous matter which occupies it) is greater than 
that furnished directly from the sun, a conclusion which we are 
not called upon here to adopt, further than to observe that 
with this, as wTith either of the other estimates, the obscure heat 
received from interplanetary matter by reflection from the sun 
at the surface of our atmosphere is considerable, so that, if our 
luminary were wholly extinguished, the temperature of the 
earth would fall much below that it would reach if only the 
direct solar heat were withdrawn. 
We shall perhaps be warranted in entertaining it then as a 
reasonable assumption, that, in the complete absence of the sun, 
the earth’s temperature would fall very nearly to —273° C. We 
prefer as the basis of an estimate, confessedly but approximate, 
to take the mean of this value and Pouillet’s, and, using the 
Fahrenheit scale, we may state that of the 500° F., which on 
the natural scale is the approximate mean temperature of our 
globe, as much as four-fifths is derived from the sun. 496 
S. P. Langley—The Solar Atmosphere. 
If it be true then that the sun is surrounded by an atmos- 
phere whose principal action in obscuring the heat radiation is 
due to a thin stratum which cuts off one-half of the heat 
which should reach us, and in whose absence this radiation 
should be doubled,—an atmosphere not independent of the in- 
terior of its globe in such a degree as our own, but one to and 
from which matter is constantly being added and withdrawn, 
it follows that any change in the ratio of supply and with- 
drawal, or other cause, which should increase its absorption by 
so much as 25 per cent would diminish the mean surface tem- 
perature of our globe by 100° F., whilst a like diminution in 
the envelope would produce a corresponding change in the op- 
posite direction. 
I am unable to see that, if the Allegheny observations are 
correct, whence I have derived the above given approximate 
value of the absorption, any other result would follow, and in 
any case the existence of life in its present forms on this planet 
seems dependent, within certain limits, on the depth and absorp- 
tive power of the solar atmosphere. The reader who may not 
have closely observed how far the validity of this result is 
dependent on, and how far independent of, the hypothesis just 
made as to the mean terrestrial temperature, is invited to 
remark that there is nothing hypothetical in any case in the 
assertion that the earth would fall, in the absence of the sun, to 
a temperature as low as any actually observed on its surface. 
If any value attach to the writer’s measurements, then, it will 
be found from a repetition of the calculations with (arctic) tem- 
peratures actually observed, and hence within the truth, that the 
result remains that a comparatively slight thickening of the 
solar atmosphere, or an increment of its absorptive power by as 
little as £, would bring the planet back to what we must sup- 
pose were the temperatures of the glacial epochs. 
Father Secchi seems to think it probable that the difference 
between the measurements of the heat at different portions of 
the solar disc made at Rome in 1852 and at Allegheny in 
1873AL are explicable by a real change in the atmosphere of 
our luminary. To the writer it seems that such a change as 
Father Secchi assumes, if real, is (if interpreted by Father 
Secchi’s own formulae) likely to have already altered the mean 
annual temperature of the earth to an amount sensible to every 
inhabitant of its surface. He is himself, therefore, of the 
opinion expressed by M. Faye that no evidence of real varia- 
tion has yet been established. 
It will not of course follow that changes may not take place 
in cycles of time long with reference to historical periods. All 
analogy leads us indeed to think an absolute uniformity most 
improbable. S. P. Langley—The Solar Atmosphere. 
497 
Finally then, it appears that though we are justified on 
grounds established by Helmholtz and Ericsson in believing in 
the constancy of the heat supply within the solar envelope 
for periods to come, which are with reference to our terrestrial 
past, almost infinite, we are not justified by our present investi- 
gations in asserting that the solar radiation has been constant 
during geologic periods, or is certain or even likely to remain 
what we now see it during corresponding periods in the future. 
If there be great cyclical changes of long period (and in our 
present ignorance we can only say that such are not antecedently 
improbable) there will be corresponding changes in terrestrial 
temperature, and it is allowable to inquire whether we do not 
find here matter for consideration in connection with those 
great changes of temperature in past epochs which have in them 
nothing hypothetical, which geology assures us have indeed 
existed, and for whose possible cause no satisfactory suggestion 
has hitherto been made. 
Errata.—Page 494, for “neucleus” read “nucleus,’1 and for “neuclei ” read 
“ nuclei.”