U S. Dept, of Agii., Bui. 63, Office of Expt. Stations 
Frontispiece. 
General View of the Atwater-Rosa Respiration Calorimeter. 
Respiration chamber in center of background. Refrigerating tank in center of foreground. Water meter on the right. Observer's table on the left. Bulletin No. 63. 
U. S. DEPARTMENT OF AGRICULTURE, 
OFFICE OF EXPERIMENT STATIONS. 
DESCRIPTION 
OF 
A NEW RESPIRATION CALORIMETER 
AND 
EXPERIMENTS ON THE CONSERVATION OF 
ENERGY IN THE HUMAN BODY. 
BY 
W. O. ATWATER, Ph. D., 
Professor of Chemistry, Wesleyan University, 
AND 
E. B. ROSA, Ph. D., 
Professor of Physics, Wesleyan University. 
WASHINGTON: 
GOVERNMENT PRINTING OFFICE. 
1899. LEITER OF TRANSMITTAL 
U. S. Department of Agriculture, 
Office of Experiment Stations, 
Washington, I). C., March 2, 1899. 
Sir: I have the honor to transmit herewith a report describing a 
respiration calorimeter of special construction and experiments with 
this apparatus on the metabolism of matter and energy and the con- 
servation of energy in the human body. The investigations described 
in this bulletin were made at Wesleyan University, Middletown, Conn., 
by W. O. Atwater, professor of chemistry at the university and special 
agent in charge of the nutrition investigations of this Department, 
and E. B. Rosa, professor of physics at the university. The expenses 
entailed in conducting these investigations have been defrayed from 
funds for nutrition investigations of this Office and from funds con- 
tributed by the Storrs Experiment Station and Wesleyan University. 
In a previous bulletin of this Office (No. 44) that portion of the 
respiration calorimeter which had to do more especially with the meas- 
urement of income and outgo of matter was described and experiments 
with man were reported, together with a brief history of the investiga- 
tions. The present bulletin describes in detail that portion of the 
apparatus which has to do with the measurements of the income and 
outgo of energy. Experimental methods are described, as well as 
check experiments, in which the accuracy of the apparatus was tested 
with heat generated by an electric current and by burning alcohol in 
the respiration chamber. Two experiments with man in which the 
complete balance of income and outgo of matter and energy was deter- 
mined are also reported. These investigations have extended over a 
number of years, but the results obtained are believed to amply com- 
pensate for the money and labor expended. An apparatus has been 
devised of such size that a man may spend a number of days in com- 
parative comfort in the respiration chamber. Notwithstanding its 
size, the results obtained in the measurement of carbon dioxid, water, 
and heat are as accurate as those secured in investigations on a much 
smaller scale in which the usual laboratory methods are followed. 
Particular mention should be made of the valuable assistance ren- 
dered by Messrs. A. W. Smith, A. P. Bryant, O. F. Tower, O. S. 
Blakeslee, and others, and especially the cooperation of Dr. F. G. 
Benedict, who has shared in the elaboration of the apparatus and 
methods as well as in the charge of the experiments. 
This report is respectfully submitted with the recommendation that 
it be published as Bulletin No. 63 of this Office. 
A. C. True, 
Director. 
Hon. James Wilson, 
Secretary of Agriculture. 
3 CONTENTS. 
Page. 
Introduction 7 
Metabolism of matter and energy 9 
Apparatus and general plan of the experiments 12 
Description of the apparatus 14 
Arrangements for preventing a gain or loss of heat through the walls of 
the calorimeter 16 
Construction of the calorimeter 16 
The thermo-electric elements 17 
Heating and cooling devices 19 
The observer's table 20 
Apparatus and method for measuring the heat 21 
The water system 21 
The water meter 24 
Determining the increase of temperature of the water 25 
Determining the temperature of the calorimeter 26 
The ventilating air current . 27 
Measuring and regulating the temperature of the ingoing air 27 
Refrigerating apparatus 29 
Meter pump for regulating, measuring, and sampling the ventilating 
air current 31 
The aspirators 34 
Arrangements for sampling and analyzing the ventilating air current. 35 
Residual carbon dioxid and water vapor in the respiration chamber 37 
Tests of the accuracy of the apparatus as a calorimeter 38 
Electrical test experiments 39 
Alcohol test experiments 50 
Composition of alcohol used-specific gravity 51 
Heat of combustion of alcohol-determination with the bomb calo- 
rimeter 52 
The calorie here used as the unit of measure 55 
The latent heat of water vapor 57 
Arrangements for the complete combustion of alcohol in the test exper- 
iment 60 
Description of the test experiments 60 
Summary of test experiments 73 
Experiments with a man 74 
Plan of the experiments 74 
Explanation of the tables 77 
Summary of results of experiments with a man 85 
General summary 87 
5 ILLUSTRATIONS. 
Pairs 
Plate I. General view of the calorimeter Frontispiece. 
II. The freezing tank and meter pump 14 
III. Ground plan of respiration calorimeter room and accessories.... 15 
IV. Horizontal section of the respiration calorimeter 16 
V. Vertical section of the respiration calorimeter 17 
VI. Diagram of electric apparatus at the observer's table 26 
VII. Blakeslee meter pump 32 
VIII. Aspirators 34 
PLATES. 
TEXT FIGURES. 
Fig. 1. Thermo-electric junctions 18 
2. Interior of respiration chamber 21 
3. Section through walls of the calorimeter, showing thermometers 22 
4. Apparatus for cooling incoming current of water 23 
5. Water meter 24 
6. Diagram illustrating use of the copper thermometer 25 
7. Coil of copper thermometer 27 
8. Air temperature regulator 28 
9. "Vestibule" 29 
10. Refrigerating apparatus 30 
11. Condensing cylinder of the refrigerating apparatus 31 
12. Arrangement of electrical apparatus for generating heat 39 
6 DESCRIPTION OF A NEW RESPIRATION CALORIMETER AND EXPERIMENTS 
ON THE CONSERVATION OF ENERGY IN THE HUMAITBODY. 
INTRODUCTION. 
The purpose of this report is to describe a new respiration calorimeter 
aud the test experiments made upon it, together with certain experiments 
made with it upon the conservation of energy in the human body.1 The 
apparatus has been devised and the methods of experimenting have 
been elaborated for use in inquiries upon the transformation of matter 
and energy in the living organism. The ultimate purpose is the study 
of some of the fundamental laws of animal nutrition. The more imme- 
diate object, however, has been the study of the transformations of 
energy. This especial study is desirable for two purposes. One is the 
demonstration, if such be possible, that the law of the conservation of 
energy holds in the living organism. The other is the practical appli- 
cation of this law in gaining more definite knowledge of the ways in 
which the body is nourished and of the value and use of food. Prefer- 
ence has been given at the outset to the problem of the conservation 
of energy, not only because of its profound scientific interest, but also 
because of its fundamental importance in the study of the laws of 
nutrition. 
In entering upon this research, to which the labor of many years must 
be given if results of the highest and most permanent value are to be 
sought, it was felt necessary to devise forms of apparatus and to elabo- 
rate methods of experimenting which should yield more detailed and 
accurate data than have been obtained hitherto. 
That there is a lack of such data is clearly shown in a recent compila- 
tion of results obtained in one of the numerous lines of modern inquiry 
'An account of preliminary experiments with this apparatus was given in Bulletin 
No. 44 of this Office entitled, Report of Preliminary Investigations on the Metabolism 
of Nitrogen and Carbon in the Human Organism with a Respiration Calorimeter of 
Special Construction, by W. O. Atwater, C. D. Woods, and F. G. Benedict. In this 
publication, however, no account was given of those parts of the apparatus and 
experiments which have to do with the measurement of the income and outgo of 
energy, so that the experiments as reported were simply so-called respiration experi- 
ments in which the balance of nitrogen and carbon was determined. A preliminary 
account of the apparatus as a calorimeter, and of the experiments described in the 
present article, was given by the writers in the Connecticut Storrs Sta. Rpt. 1897, 
pp. 212-242. 
7 8 
regarding the nutrition of man and domestic animals in health and 
disease, namely, in experiments of the class in which the balance of 
income and outgo of one or more elements was determined.1 This 
compilation includes the principal data of 3,661 experiments, of which 
2,29!) were made with man, 383 with cattle and horses, 928 with sheep, 
dogs, and other domestic quadrupeds, and 51 with poultry and doves. 
In 2,234 tests with man and 1,156 with animals the nitrogen balance 
was determiffed, and in 65 with man and 206 with animals the balance 
of carbon was determined. In 8 (with dogs), determinations were 
made of the balance of nitrogen and energy. In none was there a 
complete balance of the income and outgo of matter and energy. The 
earliest of these experiments was made in 1839; the large majority 
have been carried out within the last two decades or thereabouts. A 
few were made in the United States; a considerable number come 
from England, France, Switzerland, Italy, Scandinavia, and Russia; 
but more are from Germany than from any other country. The prob- 
lems thus studied are extremely varied. In many cases the balance of 
income and outgo was determined in connection with investigations of 
experimental methods. In a large number of cases special questions 
were investigated, the balance of income and outgo being determined 
as a means of proof of the special problem, or for some similar reason. 
Numerous experiments have been made for the purpose of studying 
the source of muscular energy in the animal body, and other theoretical 
problems. That a vast amount of valuable information on these sub- 
jects is available can not be questioned. When, however, the inves- 
tigations which have to do more especially with theoretical problems 
are examined in detail it is evident that the conclusions reached often 
contain more of partially attested theory than of well-established fact. 
Recent progress in physics and chemistry has furnished improved 
means of carrying on such investigations as those alluded to, and has 
suggested additional lines of research. It is noticeable that there is a 
tendency to go over again the ground covered by some of the earlier 
investigations and to test the theories with the aid of improved meth- 
ods. It is of the utmost importance, however, that investigators 
should cooperate in their work in order that unnecessary duplication 
may be avoided and that the best results may be obtained. Coopera- 
tion may be as successfully applied to scientific investigation as it is 
to business enterprises. This fact has been recognized in the organiza- 
tion of the nutrition investigations which are being carried out in dif- 
ferent parts of the country under the auspices of this Department, 
with the cooperation of universities, colleges, experiment stations, 
social settlements, and other institutions. It applies no less truly, 
though in a somewhat different way, to the narrower study of the 
'A Digest of Metabolism Experiments in which the Balance of Income and Outgo 
was Determined, by W. O. Atwater and C. F. Langworthy, U. S. Dept. Agr., Office of 
Experiment Stations Bui. 45. 9 
laws of animal nutrition. Not a few lines of research in this field 
have reached the point where even the best efforts of one or two indi- 
viduals can not suffice, and an extensive combination of brain work, 
hand work, and material appliances is essential for obtaining the defi- 
nite knowledge that is needed. To insure the best results experi- 
mental inquiries must be planned more philosophically and carried out 
with more thoroughness and in more detail than has generally been 
the case or, for that matter, has been possible hitherto in this country 
or abroad. This means that the research will be laborious and costly, 
but it will bring correspondingly valuable results. These considera- 
tions have been taken into account in planning the inquiry of which 
some of the results are described in this article. It will be proper to 
speak briefly of the chemical and physical principles involved in this 
special branch of physiological research before describing in detail 
the apparatus, the methods of experimenting, and the experiments 
themselves. 
METABOLISM OF MATTER AND ENERGY. 
In so far as its material phenomena are concerned, life consists in 
transformations of matter and energy. To these transformations the 
term metabolism is commonly and appropriately applied. The pro- 
cesses of metabolism are thus of two definite and closely allied kinds- 
the metabolism of matter, in which the changes are chemical, and that 
of energy, in which the changes are physical. It is commonly assumed 
that these two classes of change in living organisms, animal and vege- 
table, conform to the two fundamental laws of the conservation of 
matter and the conservation of energy.1 To say that the chemical 
transformations in the body obey the first of these laws is simply to say 
that the body can neither create nor destroy matter, a principle which 
would hardly be disputed, since it is so abundantly confirmed in the 
inorganic world and there is no reason to question its application in 
animals and plants. In like manner it might be asserted that no living 
thing, animal or plant, can either create or destroy energy, and that 
whatever energy it receives it must either retain or give off in the same 
or other form without either increase or decrease; but this has not been 
demonstrated. 
The principle of the conservation of matter in the animal organism 
has received experimental proof. While perhaps it was never seriously 
doubted, there have been discrepancies in the measurements of income 
and outgo, but improvements in experimental methods have greatly 
reduced such discrepancies. 
With the conservation of energy in the living organism the case is 
different. The methods for determining the potential energy of organic 
1 Exactly this form of statement is not usual, but it seems simple and rational and 
certainly accords with the trend of later experimental research. See references to 
the subject in U. S. Dept. Agr., Office of Experiment Stations Bui. 21, p. 99. 10 
substances are new. It is scarcely ten years since the bomb calorim- 
eter, the first satisfactory apparatus for the purpose, was perfected, and 
its use in laboratories is not yet general. Comparatively few deter- 
minations of the heat of combustion of organic substances have been 
made. Even with the aid of such determinations the conservation of 
energy is not readily proven. We can analyze a mineral or vegetable 
substance and show that the weight of the elements or compounds 
obtained exactly equals the weight of the material used for the analysis, 
and thus demonstrate that in the chemical transformations the mass 
of matter has remained the same. But when we burn a substance in 
the calorimeter and measure the heat of combustion we can do no more 
than assume that the kinetic energy obtained as heat is equivalent to 
the potential energy originally belonging to the substance. We have 
no direct means of measuring initial potential energy and comparing 
the amount with that of the kinetic energy into which it has been 
transformed. The correctness of the conclusion that the kinetic energy 
found as heat is equivalent to the original potential energy of the 
substance depends not only upon the accuracy of the determination, 
but upon the assumption that the energy has been transformed with- 
out gain or loss-an assumption which we do not doubt, but which is 
incapable of absolute proof by any method now at our disposal. These 
difficulties, however, are not so serious as they appear. If the methods 
of determining the heats of combustion of organic substances are 
new, they seem to be reasonably accurate. If they are founded upon 
assumption, the foundation seems perfectly safe. If the results thus far 
obtained are few, they already serve their purpose and they are rapidly 
increasing in number. 
The main difficulty in demonstrating that the law of the conservation 
of energy holds in the living body is in measuring the transformations 
of energy which take place therein. To prove that the law holds good 
we must know how much energy the body receives, how much it gives 
out, and how much it gains or loses of its accumulated store during a 
given time. In an experiment with a man, for instance, if we wish to 
demonstrate that the law of the conservation of energy obtains we 
must show that the income and outgo are equal. To show this equilib- 
rium of energy we must know (1) the income of potential energy which 
tlie body receives in the food; (2) how much potential energy it gives 
off in the unburned material which it excretes (in urine, feces, etc.), and 
how much kinetic energy in the heat radiated from the body, and in 
the external muscular work done; and (3) the increase or decrease in 
the store of energy in the body during the experiment. If the body is 
warmer, or if it has a larger store of proteids or fat or glycogen or other 
substance containing potential energy at the end than at the begin- 
ning, it has stored energy during the period. If, on the other hand, it 
is colder, or if its store of substance containing potential energy is less 
at the end of the experiment than at the beginning, it has lost energy. 11 
This gain or loss of energy by the body itself, the increase or diminu- 
tion of its store during the experimental period, must be measured and 
allowed for if the experiment is to be accurate and reliable. 
But even when we have measured accurately the income and outgo 
and the change of the store of energy as thus stated, we are not abso- 
lutely certain of our demonstration, although it is generally assumed 
that these factors are sufficient. It is not impossible that there may 
be other forms of energy in the outgo besides the heat radiated or 
otherwise given off from the body and the mechanical energy of exter- 
nal muscular work. It may be that the mental and nervous activities 
involve the expenditure of some form of physical energy which escapes 
our measurements. 
Previous to the time of the experiments here reported no attempt 
had been made, so far as the writers are aware, to measure the total 
income and outgo of energy in the body of an animal or a man in such 
way as either to demonstrate the application of the law of the conserva- 
tion of energy, or to show that it does not obtain. Partial measure- 
ments of the income and outgo of energy have been made from time 
to time, notably within the last five years. Such, for instance, are 
those of the experiments of Rubner1 and the later ones of Studenski 
and other Russian investigators.2 
The views of specialists as to whether the law of the conservation of 
energy actually applies in the living organism are somewhat conflict- 
ing. So far as the writers can judge, the larger number of chemists, 
physicists, and physiologists who have at all carefully considered the 
subject assume that the law does obtain, basing this supposition on 
the a priori ground that there is every reason to believe that it must 
hold in the organic world, as it has already been demonstrated to hold 
in the inorganic world. Not a few regard the experiments already 
made, notably those of Rubner just referred to, as implying very 
strongly, even if they do not strictly demonstrate, the application of 
this law in the animal body. Others, however, question this demon- 
stration, and there are some physiologists, who, knowing from long 
experience the difficulties inherent in this kind of experimenting, the 
many sources of uncertainty and error, and the great amount of labor 
which is needed for reliable results, frankly avow their belief in the 
impracticability of any satisfactory proof that the law of the conser- 
vation of energy holds in the living organism. 
These statements show the need of investigations which shall fur- 
nish experimental proof of the application of the law of the conserva- 
tion of energy to the animal body. The experiments which have been 
undertaken for this purpose by the cooperation of the Department of 
1 Ztschr. Biol., 30, pp. 119-138. See also U. S. Dept. Agr., Office of Experiment 
Stations Bui. 45, pp. 417-420. 
2 These Russian investigations are very briefly summarized in U. S. Dept. Agr., 
Office of Experiment Stations Bui. 45. 12 
Agriculture, the Storrs Station and Wesleyan University, have in some 
respects followed the lines of earlier investigations. In other respects 
they have followed new lines. It has been found necessary to develop 
methods, devise apparatus, and make many investigations before the 
respiration calorimeter was perfected and the principal problem could 
be studied. It is not at all surprising that a period of more than five 
years should have elapsed between the beginning of the actual work 
and the time when the first results were ripe for publication. It is not 
claimed that the results here reported suffice for the complete verifica- 
tion of the law referred to. They do, however, in the judgment of the 
authors, render it extremely probable. 
APPARATUS AND GENERAL PLAN OF THE EXPERIMENTS. 
Ill order to determine the total income and outgo of matter the food, 
urine, feces, and gaseous excretory products must be measured and 
their composition ascertained. In experiments of the nature of those 
reported in this bulletin, it is usual to measure the income and outgo 
of matter in terms of the elements making up the food and excretory 
products. The income and outgo of energy may be most conveniently 
expressed in terms of heat. It is necessary to know the heat of com- 
bustion of the food and excretory products, the heat of combustion of 
the material gained or lost by the body, the heat radiated from the 
body, and the heat equivalent of the external muscular work performed. 
The following statement shows the determinations which were actually 
made in obtaining the balance of income and outgo of matter and 
energy: 
INCOME. 
Matter.-Food and drink: Elements, C, H, N; compounds,water, protein (N X 6.25); 
fats (ether extract); carbohydrates (estimated by difference); mineral salts (ash). 
Energy.-Potential energy of food and drink as shown by heats of combustion; 
heat of food and drink as inferred from temperature and specific heats. 
OUTGO. 
Matter.-Feces: Same elements and compounds as in the food and drink. Urine: 
Elements, as in food and drink; compounds, water, and in some instances urea and 
other substances. Respiratory products: Elements C,H; compounds, carbon dioxid 
and water. 
Energy.-Potential energy of excretory products as shown by heats of combustion; 
heat and heat equivalent of external muscular work. 
For an ideal experiment other determinations would be desirable, but 
it is believed that those actually made are sufficient.1 
The determination of the above factors requires a special apparatus, 
capable of measuring the inspired and respired air and the heat pro- 
duced by the body. This has been called a respiration calorimeter. 
The name "respiration calorimeter" is suggested by the fact that it 
1 This ^ubject is discussed at considerable length in U. S. Dept. Agr., Office of 
Experiment Stations Bui. 44, pp. 7, 8. 13 
is essentially a respiration apparatus with appliances for calorimetric 
measurements. As a respiration apparatus it is similar in principle 
to that of Pettenkofer. As an instrument for measuring heat it is 
essentially a constant temperature calorimeter. The heat is absorbed 
and carried away by a current of water as rapidly as it is generated 
in the chamber. It is therefore a water calorimeter. The arrange- 
ments for the measurement of both the respiratory products and the 
heat given off from the body differ in important respects from those of 
any other apparatus with which we are familiar. 
The essential features of the apparatus are: 
(1) A chamber in which the subject of the experiment, a man, lives, 
eats, drinks, sleeps, and works during a period of several days and 
nights. The chamber is furnished with a folding chair, table, and bed. 
(2) Arrangements for ventilation by a current of air which is drawn 
from out of doors and passes through the chamber. The volume of this 
air current is measured and its temperature is so regulated as to be the 
same on entering the chamber as upon leaving. Samples for analysis, 
taken before it enters and after it leaves the chamber, give data show- 
ing the amount of carbon dioxid and water given off from the body 
through the lungs and skin. 
(3) Arrangements for passing food and drink into the chamber and 
removing the solid and liquid excreta. These materials are analyzed 
and give data for calculating the income and outgo of nitrogen, and of 
sulphur, phosphorus, chlorin, and metallic elements if these are deter- 
mined. Taken in connection with the determinations of carbon dioxid 
and water in the respiratory products, they show the income and outgo 
of carbon and hydrogen. The analyses of the food, and of the solid and 
liquid excreta, enable us to determine the so-called digestibility of the 
food, i. e., the proportions of nutrients actually made available. 
(4) Arrangements for measuring the heat given off from the body of 
the man in the chamber and the heat equivalent of the external mus- 
cular work. The heat given off is carried away by a current of cold 
water which passes through a series of pipes inside the chamber. By 
regulating the temperature of this water current as it enters, and also 
its rate of flow, it is possible to carry away the heat just as fast as it is 
generated, and thus maintain a constant temperature inside the cham- 
ber. The amount of the outgoing water and its increase in temperature 
are measured, thus determining the amount of heat carried away. 
In order that the heat taken up by the absorbers and carried out by 
the water current shall represent exactly the amount given off' from 
the man's body or otherwise produced in the chamber, it is necessary 
to provide that there shall be no passage of heat through the walls, or 
rather, that the small quantities that may pass in and out shall exactly 
counterbalance each other, and that the ventilating current of air 
shall leave the chamber at the same temperature as it enters, so that it 
shall carry out neither more nor less heat than it brings in. The excess 14 
of water vapor in the air leaving the chamber over that in the air enter- 
ing represents water given off from the body of the subject, and the 
heat required to vaporize it must be added to the heat carried off by 
the current of water to obtain a measure of the total heat given off' by 
the subject. 
In the actual experiments the man remains for several days and 
nights (generally four days and five nights) in the chamber. The food 
and drink are passed in and the feces and urine removed at regular 
intervals. The diet is decided by the special question which is being 
studied, but is made agreeable to the subject of the experiment. 
In some of the experiments the man had as little muscular exercise 
as practicable. In these so-called "rest experiments" no attempt was 
made to measure the muscular work. In others, so-called " work experi- 
ments," the amount of work done was measured by a specially devised 
ergometer, consisting of a stationary bicycle which was belted to a 
small dynamo. The electric current generated passed through an 
incandescent lamp inside the chamber, where its energy was trans- 
formed into heat. The strength of current and voltage was deter- 
mined by instruments outside the calorimeter. The heat equivalent of 
the muscular work done was thus added to and measured with the 
heat given off from the body. The duration of the work and the 
amount of the electric current generated gave data for the computa- 
tion of the amount of work performed. 
The men who have served as subjects of the experiments have not 
found their sojourn in the chamber disagreeable nor have they suffered 
any special inconvenience. The experiments are, however, very labo- 
rious, and require not only the services of trained observers day and 
night for the chemical and physical determinations, but also a con- 
siderable amount of painstaking work in preparing the food, taking 
care of the excreta, and attending to the numerous details before, 
during, and after the experiment. 
DESCRIPTION OF THE APPARATUS. 
A general idea of the apparatus can be had from Pl. I (frontispiece) 
taken from a photograph. This gives a general view of the princi- 
pal parts of the apparatus, which stands in a basement room of the 
Orange Judd Hall of Natural Science, at Wesleyan University. In the 
center is the large chamber, which is surrounded by a sheathing of 
wood. At the end of the chamber, on the right, is shown a door, which 
serves also as a window. At the right of the window and just below 
it are the arrangements for cooling and measuring the current of water 
which brings away the heat from the interior of the chamber. At the 
left, in front of the large brick pillar, is a table 1 at which an observer 
sits to record the temperature of the interior of the apparatus and of 
'This table has lately been moved to a position just in front of the apparatus. This 
in no wise modifies the method of experimenting. U. S. Dept, of Agri., Bui. 63, Office of Expt. Stations. 
Plate II. 
The Freezing Tank and Meter Pump. U. S Dept, of Agri., Bui. 63, Office of Expt. Stations. 
Plate III. 
BALANCES 
METER 
PUMP 
Ground Plan of Respiration Calorimeter Room with Apparatus and Accessories. 
THREE ASPIRATORS 
RESPIRATION CHAMBER 
OBSERVERS 
TABLE 
FAN BOXES 
AMMONIA 
REFRIGERATING 
MACHINE 
DRYING OVENS 15 
the currents of air and water, the temperature being measured electric- 
ally. Behind the brick pillar is an ammonia refrigerating machine, not 
shown in the picture. Its purpose is to cool the solution of calcium 
chlorid which is contained in a large tank in the center foreground. 
This tank (Pl. II, and tig. 10, p. 30) is inclosed within a double walled 
wooden casing. The liquid ammonia is evaporated in the coiled iron 
pipe, shown in fig. 10, p. 30. The ventilating current of air, before it 
enters the chamber, is passed through the coppercon denser A, which is 
immersed in the calcium chlorid solution in the right-hand corner of 
this tank, and thus cooled to a temperature of from -19° to -22° C. 
At this very low temperature nearly all of the water is removed from 
the air, so that it enters the chamber quite dry. Just before entering 
at the right of the glass door it is warmed to the temperature of the 
interior of the chamber. On coming out it passes through a second 
condenser, consisting of a pair of copper cylinders B, C, in the cold brine 
in the tank, and thus the larger part of the water which has been 
imparted to it by the respiration of the man inside the chamber is frozen 
and deposited. By means of valves contained in the valve boxes above 
the cylinders, the air current may be diverted to a second pair of cooling- 
cylinders D, E. The first pair may then be removed and the deposited 
moisture determined by the increase in weight. The air pump is shown 
at the right of Pl. II. 
The disposition of the various parts of the apparatus is further shown 
in the ground plan of the room (Pl. III). On the north and east sides 
of the room the massive stone walls of the building appear in section, 
interrupted by five large windows. The window of the respiration 
chamber opens toward the east, so that the interior of the chamber is 
lighted from the two windows in the northeast corner of the room. At 
the west end of the room stands a 2-horsepower electric motor, belted 
to the driving pulley of a line shaft extending the entire length of the 
room. From this shaft power is taken to drive the ammonia refrigerat- 
ing machine, the fans contained in the two fan boxes which circulate the 
air in the space surrounding the respiration chamber, and the meter 
pump which propels and measures the ventilating air current. A 
compression air pump is also driven from the shaft. This provides the 
compressed air which operates the valves of the meter pump. The air 
current passes from out of doors through the pipe Pi into the refrig- 
erating tank where the greater portion of its moisture is deposited, and 
thence by the pipe P2 to the respiration chamber (fig. 2), after being 
warmed to the temperature of the latter. From the respiration cham- 
ber the air passes through the pipe P3 to the refrigerating tank, where 
its moisture is again deposited for the most part, and thence to the 
meter pump through P4. Here the pump records the volume of air and 
delivers every fiftieth cylinderful to the pans for analysis for carbon 
dioxid and residual moisture. Through small pipes S3, S4, samples for 
analysis of the dried air of the pipes P2, P4 are drawn by the aspirators 
shown on the south side of the room; and through the pipes Si and 16 
S2 other similar samples are drawn off by aspirators in an adjoining 
room for duplicate analyses. The details of these operations will be 
given further on. On the north side of the pier near the middle of 
the room, and attached to the pier, is the galvanometer, used for meas 
uring electrically temperatures and differences of temperature; and on 
the observer's tables are located Wheatstone bridges, keys, and resist- 
ances used in the measurement and regulation of temperatures, as 
hereafter described. Tables Tb T2, T3 and other appliances afford 
facilities for part of the chemical work done in connection with the 
experiment, but most of the chemical work is done in other portions of 
the building. 
ARRANGEMENTS FOR PREVENTING A GAIN OR LOSS OF HEAT 
THROUGH THE WALLS OF THE CALORIMETER. 
CONSTRUCTION OF THE CALORIMETER. 
The construction of the calorimeter is shown in horizontal section in 
Pl. IV. The inner chamber is practically an apartment with double 
walls of metal. The inner of the two metal walls is of sheet copper, and 
incloses a space 2.15 metei^ (7 feet) long, 1.92 meters (6 feet 4 inches) 
high, and 1.22 meters (4 feet) wide, the corners being rounded. Copper 
was chosen because it can be obtained in large sheets, can take a high 
polish and so the more fully reflect radiant energy, is a good conductor 
of heat and thereby tends to equalize local differences of temperature, 
and has a rather small specific heat. A window in one end, 70 centi- 
meters high by 49 centimeters wide, serves also the purpose of a door. 
During an experiment, however, the door is sealed and a circular open- 
ing; E, which has tightly fitting caps a, b, both inside and outside, is 
used for communication with the interior. Outside is a box, c, filled 
with nonconducting material to prevent the passage of heat through 
E. The seams are soldered and the chamber is air-tight when the door 
and other openings are sealed. The outer metal wall is of zinc, which 
is less expensive than copper, and serves the purpose sufficiently well. 
Between the two walls is an air space, A, of 7.6 centimeters (3 inches). 
In this stands a wooden framework to which the two metal walls are 
securely attached. 
In order to protect the calorimeter from the fluctuations of tempera- 
ture of the basement room in which it stands it is inclosed within 
three concentric walls of wood. Through each of these is an opening 
corresponding to the window in the wall of the calorimeter proper. 
These openings are closed by tightly fitting glass doors, hung on hinges. 
Near these doors are shown the pipes through which the water enters 
and leaves the calorimeter. Its temperature on entering is measured 
by the mercury thermometer, G; and as it leaves, by a similar ther- 
mometer, H. Between the zinc wall and the innermost wooden wall is 
a 5 centimeter (2-inch) air space, B; between this wall and the next is 
a third air space, C, of 5 centimeters (2 inches); and, finally, between U. S. Dept, of Agri., Bui. 63, Office of Expt. Stations. 
Plate IV. 
Horizontal Section of the Respiration Calorimeter. 
SCALE I METER 
OUTER REGULATING CURRENT 
DEAD AIR 
SHADED THUS 
ARE OF WOOD. 
THE PORTIONS U. S. Dept, of Agri., Bui. 63, Office of Expt. Stations. 
Plate V. 
Vertical Section of the Respiration Calorimeter. 17 
this wall and the outer one is a fourth air space, D, of 5 centimeters (2 
inches). The wooden walls are made of matched pine, covered with 
building paper, and the outer one is double, with paper between. Of 
these four air spaces A and C are "dead" air spaces, while the air in 
the spaces B and D can be kept in constant circulation by means of 
rotary fans driven by power. Each of the air spaces B and D is con- 
tinuous around the sides and over the top and bottom, and each com- 
municates with its fan box in the rear by means of one passage extend- 
ing from the top of the air space to the top of its fan box, and another 
from the bottom of the air space to the bottom of its fan box. Thus 
the air may be kept in constant circulation through these two air 
spaces, and by heating or cooling this moving air the temperature of 
the corresponding space may be raised or lowered. We may also look 
upon the moving strata of air as shields, guarding the interior space 
occupied by the calorimeter from changes in temperature without. A 
vertical section of the calorimeter is shown in Pl. V. The walls at the 
top and bottom are constructed like those of the sides, but as this sec- 
tion is taken through the center of the chamber it shows the flues for 
the regulating air currents. The air from the space B is drawn into 
the upper part of the bottom flue, as indicated by the arrows, and by 
means of the inner fan is driven into the lower part of the flue at the 
top. Openings along this flue allow the air to flow out over the top of 
the calorimeter, and thence down around the four sides to the bottom 
and so back to the fan again. 
Similarly the air in the space I) is kept in circulation by the fan in 
the outer fan box. The circulation in this space is in the opposite 
direction from that in B; that is, the air is taken from the top and 
driven in at the bottom. As the upper part of the room is usually 
much warmer than the floor, this arrangement for the outer circulating' 
air tends to equalize the«difl'erence of temperature by drawing off the 
hot air from the top and driving it in at the bottom. At the front are 
shown the glass window and doors as closed. 
THE THERMO-ELECTRIC ELEMENTS 
In order to determine whether the inner or outer metal wal1 of the 
calorimeter is the warmer-that is, in order to know whether to heat or 
to cool the air space B-we use a large number of pairs (304) of thermo- 
electric junctions, distributed over the four sides, top, and bottom of 
the calorimeter, one-half of the junctions (the iron-German-silver) being 
in close thermal contact with the copper wall and the other half (the 
German-silver-iron) with the zinc wall. One of these elements is shown 
in fig. 1, where Ci is a copper cartridge shell or cap, soldered to the 
copper wall, and maintaining the inclosed thermo-electric junction at 
the same temperature as the copper wall, and C2 is a similar cap soldered 
to the zinc wall. There being copper leads connecting the several pairs, 
17951-No. 63 2 18 
there are two junctions (copper-iron and German-silver-copper) in the 
outer cap; but these are thermo-electrically equivalent to one German- 
silver-iron junction. In reality there are four pairs of junctions in each 
group instead of a single pair as shown in the figure. These thermo- 
electric elements are made of silk-covered iron and German-silver wire 
No. 20. They are all in series, and are joined to a sensitive!I'Arson val 
galvanometer. The galvanometer is quick and deadbeat, and readings 
may be taken rapidly. If the copper wall is on the whole warmer than 
the zinc, the electromotive forces due to the junctions in contact with it 
overpower those due to the junctions in contact with the zinc wall and 
a current flows through the galvanometer, proportional in amount to 
the difference of temperature of the walls, and produces a deflection 
to the right. It is then 
necessary to warm the 
space B, which raises 
the temperature of the 
zinc wall, and the de- 
flection will be reduced. 
If the deflection is to 
the left, the space B is 
cooled until the deflec- 
tion is zero. A deflec- 
tion of 1 division on 
the scale represents a 
difference in tempera- 
ture between copper 
and zinc of about 
0.007° C. An experi- 
enced observer can or- 
dinarily keep the de- 
flection within one or 
two divisions, and by 
taking frequent read- 
ings at regular inter- 
vals the + and - deflections are separately summed up and made to 
cancel each other as the experiment proceeds, and hence no "cooling 
correction" is necessary. 
Of course the deflection of the galvanometer is due to the algebraic 
sum of all the electromotive forces at the 304 pairs of elements. At 
some places the copper maybe warmer and at other places colder than 
the zinc, and the galvanometer indicates an average difference. In 
order that this average may be the more reliable the elements have 
been spaced as nearly uniformly as possible over the entire surface. 
There are 76 groups of elements of 4 pairs each: Of these 76 groups 
11 are on the top, 11 on the bottom, 22 on the upper part of the four 
sides (over two-fifths of the whole areni, and 32 on the lower part of 
the four sides (covering the remaining three-fifths of the area). Thus 
■COPPER WALL 
■ZINC WALL 
- COPPER 
COPPER 
IRON 
GERMAN SILVER 
Fig. 1.-Thermo-electric junctions. Ct and C2 are ordinary cop- 
per cartridge shells soldered to the copper and zinc walls,* 
respectively, and contain the junctions of the thermo-electric 
elements. Only one pair of junctions is shown in the figure, but 
there are really four such pairs in each group. The junctions 
are electrically insulated from the walls of the calorimeter, but 
in the best possible thermal contact. 19 
the elements are joined up in four sections, designated respectively as 
top, upper, lower, and bottom, and each section can be joined to the galva- 
nometer separately. In practice, therefore, we can determine whether 
to heat or cool any of the four sections, top, upper, lower, or bottom, in 
order that the deflection shall always be small for each section, and 
that the total deflection for all shall be zero. This secures a more per- 
fect balance than could be had if only the total deflection were zero; 
for then an appreciable quantity of heat might be flowing in at the top 
and out at the bottom and the two quantities not exactly balancing 
each other. 
There are 26 pairs of thermo-electric elements set in the wooden 
wall between the air spaces B and D, one set of junctions extending 
into the air space B and the other into D. These elements indicate 
the difference of temperature between B and D, and indicate when D 
must be heated or cooled to keep it nearly at the temperature of B. 
HEATING AND COOLING DEVICES. 
In order to be able to heat or to cool any of the four sections of the 
air space B separately, a system of German-silver wires for heating 
and of iron pipes for cooling are arranged in this space, between the 
inner wooden wall and the zinc wall. German-silver wire No. 30 was 
stretched horizontally around the calorimeter, being supported and 
insulated by passing it through glass tubes set into vertical wooden 
strips which are attached to the zinc wall. The wire was divided into 
four sections corresponding to the four sections of the thermo-electric 
elements and the ends brought out so that an electric current could 
be passed through any of them at will, the strength of the current 
being regulated by variable resistances. In a similar manner |-inch 
(6 millimeters) iron pipes are arranged in four sections so that by 
opening a valve water from the city service can be passed through 
any of the pipes, to cool the surrounding space; when the water is 
shut off, a vent in each of the valves allows the pipes to drain. When 
only slight cooling is needed, the water may be allowed to flow only a 
few seconds at a time. In this manner the temperature of the air 
space can be perfectly controlled, and when the temperature within 
the calorimeter remains constant, as when heat is being generated at 
a uniform rate, the deflection can usually be kept within 1 division, 
which means an average difference of temperature between copper and 
zinc of less than 0.01° C. 
Inasmuch as the temperature of the room is always lower near the 
floor than near the ceiling, the air circulating through the space D is 
warmed by passing over electric lamps or through a pipe heated by a 
Bunsen burner and made to enter at the bottom of the calorimeter. 
This tends to equalize the temperature between the top and the bot- 
tom, and guided by the indications of the second set of thermo electric 
junctions, the air in D is kept nearly at the temperature of that in B. 20 
Thus the space B is perfectly protected from changes of temperature 
without, and it is possible to keep its temperature almost the same as 
that within the chamber. 
THE OBSERVER'S TABLE. 
The terminals of the thermo-electric circuits and of the galvanom- 
eter are brought to the observer's table (Pl. I). The galvanometer 
stands on a shelf attached to a brick pier, about 2 meters beyond the 
table, and the galvanometer scale is directly over the table (Pl. Ill, p. 15), 
being also attached to the pier. A 4-point switch, N, joins the gal- 
vanometer to the several thermo-electric circuits in succession. If the 
key is on the first point, the thermo electric circuit of the copper- 
zinc walls, designated as No. 1, is joined to the galvanometer, and 
the deflection indicates their difference of temperature, as already 
explained. An auxiliary switch, M, throws into the circuit any of the 
four sections separately, or all together when desired. 
The 26 pairs of thermo-electric elements, designated as circuit No. 
2, set in the wooden wall between the spaces B and 1), are joined in 
series, and the terminals are brought out to the switch N, and joined 
to the galvanometer when the key is on the second contact. The gal- 
vanometer then indicates whether the space 1) is warmer or colder 
than B, and thus shows whether to heat or to cool the air flowing 
through it. If the temperature of the room is not very different from 
that of the calorimeter it is not necessary to heat or cool the air in D. 
But if the room is much colder than the respiration chamber, as is 
often the case at night when there is no steam in the radiators in the 
room, then by heating the air of 1) before it enters at the bottom it 
protects the space B from outside influence. The temperature of the 
room not infrequently falls to 15° or 13° C., while that of the respira- 
tion chamber is maintained constant, usually at about 22° C. 
When the key is on the third contact the galvanometer is short cir- 
cuited, and when on the fourth the thermo-electric junctions, designated 
as circuit No. 4, in the ventilating air current, hereafter to be described, 
are joined to the galvanometer. 
A bank of incandescent lamps, LL, is divided into five groups of 
7 lamps each. Each of four groups is in series with one section of 
the German-silver heating coils already referred to, and which are 
wound around the calorimeter in the air space B. For example, the 
first section is in series with that section of the coil distributed over 
the top of the calorimeter, and the current is regulated by varying the 
number of lamps in series. Any one or more lamps can be cut out of 
circuit by a short-circuiting key. The lamps have different resistances, 
and by suitable combinations any desired current may be obtained, 
the source of current being the 220-volt power circuit from which the 
electric motor is driven. 
The fifth section of lamps is in series with an incandescent lamp 21 
placed in the pipe through which the ventilating current of air enters 
the respiration chamber, and by regulating the current through it, the 
temperature of the air as it enters is maintained the same as that of 
the air leaving. 
Bridge No. 6 is a Wheatstone bridge, by means of which the differ- 
ence of temperature of the stream of water as it enters and leaves the 
Fig. 2.-Interior of respiration chamber. This is the view which would be secured by breaking 
away a portion of one side and end of the interior copper wall of the respiration chamber. The 
ventilating air enters the chamber through the bottom of W, leaving from the rear of the chamber 
through the long tube connected with the upper end of W. The copper tubes H, H, strung with 
copper disks I, I, constitute the absorbers through which flows the water current which carries away 
the heat generated in the chamber. J, J are the copper troughs for collecting the drip from the 
absorbers. M, M, M are electrical thermometers which show the temperature of the air in the 
chamber; N, N those which show the temperature of the copper wall. 
calorimeter is measured, and bridge No. 5 is a similar Wheatstone 
bridge for determining the temperature of the respiration chamber. 
These bridges will be described later. 
APPARATUS AND METHOD FOR MEASURING THE HEAT. 
THE WATER SYSTEM. 
Having arranged the calorimeter so that no heat can escape or enter 
through its walls, it remains to carry away and measure the heat gen- 
erated. This is done by a stream of water flowing through a copper 22 
"absorber" (HH, fig. 2). A copper pipe of 6 millimeters Q inch) bore 
was bent into a rectangle 90 centimeters by 182 centimeters (3 feet by 
6 feet). In order to increase the absorbing area of the pipe, one thou- 
sand disks II, 5 centimeters (2 inches) in diameter and with a 1 centime- 
ter (f inch) hole in the center, stamped out of sheet copper, were slipped 
over the pipe like beads on a string, and soldered 6 millimeters (| inch) 
apart. The pipe and disks were then painted black. This absorber is 
suspended about 25 centimeters (10 inches) below the ceiling of the 
calorimeter, and 15 centimeters (6 inches) from the sides and ends. 
Water enters the calorimeter from the coolers through a pipe, E (figs. 
2 and 3), containing a mercury thermometer, G (fig. 3), and after flow- 
ing through the absorber leaves through a second pipe, F, near to and 
parallel with the first, containing a second thermometer, H. These 
Fio. 3.-Section through walls of the calorimeter, showing thermometers. A is the copper and B the 
zinc wall of the calorimeter chamber. C and I) are the inclosing walls of wood and paper. The 
long thermometers G, II, extend into the two pipes E, F, through which the current of water enters 
and leaves the calorimeter. Readings are taken by means of the microscopes, the scale being illu- 
minated by small electric lamps. The rate of flow of water is regulated by a valve with an index 
moving over a graduated scale. 
two water pipes pass through holes in a cylindrical block of wood. 
The bulbs of the thermometers are at a point, g, h (fig. 3), midway 
between the zinc and the copper walls. Hence one gives the tempera- 
ture of the water just as it enters the respiration climber and the 
other its temperature as it leaves. The thermometers are 60 centime- 
ters (24 inches) in length, and are graduated from 0° to 30° C., in 
twentieths of a degree, and read to hundredths of a degree. They 
have been carefully compared with our standards. These thermome- 
ters are read by microscopes mounted as shown in fig. 3, the thread of 
mercury being illuminated by small electric lamps. Their difference 
of temperature gives a measure of the heat absorbed within the calo- 
rimeter per gram of water. 
In order to regulate the temperature of the ingoing water, and espe- 23 
cially to have it as low as required, it is passed through a cooler. This, 
as shown in fig. 4, consists of a pair of closed copper cylinders sub- 
merged in a bath of cold calcium chlorid. The temperature of this 
bath is kept at the desired point by pumping in cold calcium chlorid 
solution from the large refrigerating tank as often as necessary. The 
water to be cooled enters the first cooler, being carried to the bottom 
where the warm water prevents the formation of ice. This water being 
partially cooled enters the second cooler, being carried to the bottom 
as in the first one. Here it is cooled still further before going to the 
Fig. 4.-Apparatus for cooling incoming current of water. The two cylinders shown within the 
tank are surrounded by a solution of cold calcium chlorid drawn from the large refrigerating 
tank by means of the pump shown. 
absorbers. By this means the ingoing water can be brought as low as 
2° C. if desired, and any higher temperature can be obtained either by 
varying the temperature of the bath or by mixing the cold water with 
warmer taken directly from the supply. At the top of the room is a 
tank which is kept full and overflowing by a stream of water from the 
city service. From this tank is taken the water which flows through 
the absorbers. Having thus a constant head, the rate of flow of water 
through the absorbing pipes is uniform. A regulating val ve with a 
pointer moving over a graduated arc (fig. 3) enables one to vary the 
rate of flow within wide limits. 24 
THE WATER METER. 
The water is received in a special form of water meter (fig. 5), having 
a capacity of 10 kilograms. The water which has flowed through the 
absorbers runs into the 
small cup A, which is di- 
vided by a vertical par- 
tition through the cen- 
ter and reaching from 
the bottom about half- 
way to the top. On 
either side of this parti- 
tion are small tubes lead- 
ing from the bottom of 
the cup A to the cylin- 
ders B and C. When 
the water running into 
A falls on one side of the 
partition it will run down 
the tube to the corre- 
sponding cylinder,which 
will be filling up. As 
the level of the water 
rises it will finally be as 
high as the top of the 
partition in the little 
cup A, when no more 
water will enter, but will 
overflow and run into 
the other cylinder. The 
first cylinder now con- 
tains just 10 kilograms 
of water. 
In the narrow neck of 
the cylinder is a float, 
F, in the form of a test 
tube. This is hung from 
the end of a spring, and 
as the rising water lifts 
the float the spring 
comes in contact with 
the point H and closes 
the circuit of an electric 
alarm bell which con- 
tinues to ring until the 
siphon S is started and 
the cylinder begins to empty. After the siphon is started it will 
draw all the water from the cylinder down to a certain level without 
TO BELL AND BATTERY 
Fig. 5.-Water meter. The stream of water which carries away 
the heat generated in the respiration chamber flows into one of 
the cylinders B, C, until it is tilled, when the stream is diverted 
automatically by the apparatus A, to the other cylinder. The 
first is emptied by means of the siphon S. The siphon is 
operated by the attendant whose attention is called by an 
electric bell which begins to ring as soon as the cylinder is 
filled, and continues ringing until the siphon is started. 25 
any further attention. The stream of water is diverted to the other 
side of the cup A by the operator before the siphon is started. Record 
is made at the instant the meter is filled, and the interval between suc- 
cessive records gives the period for the corresponding 10 kilograms of 
water. Thus the product of the average difference of temperature 
between Ihe ingoing and outgoing water by 10 gives the number of 
calories of heat brought away during the given period. 
DETERMINING THE INCREASE OF TEMPERATURE OF THE WATER. 
As a check upon the thermometer readings, and in order to avoid too 
frequent readings of the thermometers, two coils of copper wire are 
sealed in thin copper tubes and inserted in the water pipes near the 
bulbs of the thermometers. They are joined to two comparison coils 
on the slide-wire Wheatstone 
bridge No. 6 (Pl. VI), graduated 
so that readings on the scale give 
differences of temperature of the 
water in the pipes, 1 millimeter 
being 0.01° C. Headings on this 
bridge are usually taken every 
two minutes during an experi- 
ment, and careful readings on 
the mercury thermometers are 
taken occasionally for calibration 
purposes. This bridge is equiva- 
lent to what is generally known as 
a platinum thermometer, but inas- 
much as copper coils are used 
instead of platinum, it is here 
called a "copper thermometer." 
The resistance of copper increases 
with its temperature, and hence 
the difference of temperature of two coils of wire may be accurately 
measured by determining their resistances. 
The principle employed for the measurement of these resistances is, 
as just stated, that of the Wheatstone bridge, illustrated in fig. 6. The 
four arms are A, B, C + c, D + d. The battery is joined at Bi and B2, 
the galvanometer at Gi and G2. If A and B represent the proportional 
arms of the bridge, which are here equal, C and D the two copper coils 
in the water pipes, and c + d the slide wire of the bridge, then for no 
current through the galvanometer the following proportions must be 
true: 
A:B::C + c:D + d. 
Suppose now the coil D is warmed so that its resistance is increased. 
To keep the bridge still balanced, B2 must be moved along the slide 
wire reducing d, and, at the same time, increasing c, until the bridge 
Battery. 
Fig. 6.-Diagram illustrating use of copper ther- 
mometer. This is a Wheatstone bridge in which 
A and B are two equal comparison coils, C is a 
fixed resistance, D is the resistance coil in the 
calorimeter, and c and d are the two parts of the 
straight wire on the bridges shown in Pl. VI. 26 
is again balanced. By properly graduating the slide-wire scale the 
position of B will show directly the difference in temperature between 
the coils C and D; and these coils of fine insulated copper wire, inclosed 
in very thin copper tubes, quickly take up the temperature of the 
water by which they are surrounded. 
.Referring now to Pl. VI, the connection of bridge No. 6 will be 
readily seen. The parts are lettered the same as in fig. 6. In addition 
are the small coils cz, czz, cz,/. The length of the bridge wire corre- 
sponds to a difference of temperature of 5° in the coils C, D. If the 
difference increases above this the coil cz is added to C, so that the 
slider instead of being at "5" will be at "0", and then differences of 
temperature between 5° C. and 10° C. can be measured. In the same 
way, when the difference reaches 10° C., c" is added to C+cz, and "0" 
then corresponds to 10° C. The connections are made and changed by 
means of a heavy link dipping into mercury cups. The scale is 500 
millimeters in length, and hence the position of the slider can easily be 
read oft' in hundredths of a degree. 
The rate at which heat is taken up by the absorbers depends upon 
their temperature as compared with that of the surrounding space and 
the freedom of circulation of air around them. Three methods are 
employed to diminish this rate below the maximum. First, the water 
may be allowed to enter at a higher temperature than the minimum of 
2° C.; second, the water may be made to flow slowly through the ab- 
sorbers so that its average temperature is higher; and third, shields 
may be put around theabsorbers. Deep copper troughs J, J, hang under 
the absorbers (see flg. 2), one on each side and one on each end. The 
side troughs or shields can be operated from without by means of cords 
passing over pulleys L, L, which are wound up by rods reaching through 
the side of the calorimeter. Raising a shield around a portion of the 
absorber impedes the circulation of the air, the colder air within the 
shield being able to escape only by rising and flowing over the sides. 
As these shields can be raised to any desired point, and the temper- 
ature and rate of flow of the entering water can be regulated at will, 
any desired rate of absorption from 1 to 250 calories per hour, its maxi- 
mum rate, can be obtained. 
DETERMINING THE TEMPERATURE OF THE CALORIMETER. 
In order to measure, not only the total amount of heat evolved by 
the subject, but also the rate at any and all times, it is necessary to 
keep the interior temperature of the calorimeter very nearly constant, 
and also to measure the small changes in temperature which may occur. 
For this purpose we have mounted six coils of No. 32 copper wire near 
the walls of the chamber, connected them in series with one another, 
and joined them to a slide-wire bridge (No. 5, Pl. VI) outside, so 
graduated as to read off the temperatures directly in degrees Centi- 
grade, 1° extending over a hundred millimeters on the scale. Thus U. S. Dept, of Agri., Bui. 63, Office of Expt. Stations. 
Plate VI. 
Diagram of Electric Apparatus at the Observer's Table. 
Bridge No. 5 is a special Wheatstone bridge used in measuring the temperature of the respiration chamber. Bridge No. (i is used to 
measure the difference of temperature of the ingoing and outcoming water. R, S are two electric keys. L, L is a bank of 
incandescent electric lamps used as variable resistance. M, N are two multiple switches. 
No 1 
BOTTOM- 
LOWER 
UPPER 
TOP 
No.4' 
No.2 
Gal. 
220 
VOLTS 
HEATING COILS IN SPACE B 
Bridge No.6 
Bridge No. 5 
No. 5 
No. 7 27 
the temperature is read to the hundredth of a degree. This bridge is 
similar to No. 6. A and B are the proportional arms of the bridge, 
here equal to each other, C is the comparison 
coil mounted on the bridge. The wire of coil C 
(as well as both bridge wires) was kindly fur- 
nished by the Weston Electric Instrument Com- 
pany, and has a zero temperature coefficient, so 
that its resistance is constant. Hence the posi. 
tion of the slider is changed only when the tem- 
perature of the copper thermometers D varies. 
At K is a switch for short circuiting the small 
coil c'. When this is closed the scale reads from 
17° C. to 22° C.; when open, from 20° C. to 25° 
C. The copper coil D, which forms the fourth 
arm of the Wheatstone bridge, is subdivided 
into six separate coils. Two of these are 
mounted on each side of the chamber and one 
on each end, each being at a different height 
M, M, M (fig. 2). Thus they indicate a fair aver- 
age temperature. These coils are wound on 
light wooden frames (fig. 7), and are covered by 
a perforated brass tube. They are thoroughly 
insulated from the copper wall, and further pro- 
tected from injury by guards O (fig. 2). They 
are very sensitive to slight changes in the tem- 
perature of the air. 
A second set of similar coils is inclosed in 
copper boxes and attached in close thermal con- 
tact to the copper walls. They may be joined to 
the same bridge by a key, S, and indicate the 
temperature of the copper walls. These copper 
coils, with their slide-wire bridge, we call the 
"copper thermometers." A mercury thermome- 
ter passes through the wooden and zinc walls, 
and its bulb lies in a copper pocket which is 
soldered to the copper wall. This gives the temperature of the copper 
at that place, and its readings are compared with the "copper ther- 
mometer" readings. 
Fig. 7.-Coil of copper ther- 
mometer. These coils are 
mounted on the inner wall 
of the respiration chamber. 
They are joined in series, 
and indicate by their vary- 
ing resistance the fluctua- 
tions in temperature of the 
calorimeter. 
THE VENTILATING AIR CURRENT. 
MEASURING AND REGULATING THE TEMPERATURE OF THE INGOING 
AIR. 
The temperature of the outcoming air will be the same as that inside 
the chamber. If the ingoing air can be kept at the same temperature 
there will be no gain or loss of heat due to the circulation of the air. 
The ingoing air passes through the "regulator" before entering the 28 
chamber. This regulator, shown in lig. 8, is a copper cylinder 76 centi- 
meters (30 inches) long and 18 centimeters (7 inches) in diameter. 
Within this is another cylinder 5 centimeters (2 inches) in diameter 
extending through the ends of the outer one. On the outside of this 
small cylinder are soldered eight vanes V, which reach nearly to the 
outside wall, thus increasing the effective area of the cylinder, which is 
tilled with water to increase its capacity for heat. The air comes to the 
regulator a few degrees too cold, and is warmed by an incandescent lamp 
L, placed in the pipe. The current in the lamp is varied to give the re- 
quired amount of heat. 
After passing over 
the lamp and being 
warmed to about the 
desired temperature 
the air enters the regu- 
lator S,where it attains 
its final uniform tem- 
perature. If it enters 
a little warmer it will 
impart its surplus heat 
to the water in the 
smaller cylinder T, 
while if it enters 
slightly cooler it will 
receive heat from the 
same source. Thus it 
will enter the chamber 
through the tube U at 
a nearly constant tem- 
perature. This regu- 
lator of large capacity 
for heat steadies the 
temperature of the air 
current as the massive 
fly wheel of an engine 
steadies its speed. 
Within the chamber 
is the "vestibule" (fig. 
9), through which pass the currents of ingoing and outcoming air. This 
is a double-walled copper chamber, divided into two portions by a double- 
walled partition. The outcoming air enters the top and passes out 
through the tube U2. This space will then have the temperature of the 
outcoming air. In the same way the ingoing air enters through the 
tube U] and leaves at the bottom. The lower space is then at the tem- 
perature of the ingoing air. Extending through the partition on either 
side are twenty-four pairs of iron-German-silver thermo electric couples. 
These are connected to the galvanometer through point 4 of switch N 
Fig. 8.-Air temperature regulator. The ventilating current of air 
passes through this regulator on its way to the calorimeter and 
the temperature is raised by the lamp L, or lowered by water 
circulating in the cylinder T, as may be necessary to make it 
the same as the temperature of the air escaping through the exit 
pipe U. 29 
(Pl. VI), and if there is any difference of temperature between the air 
in the two spaces it will cause a deflection of the galvanometer. If 
this shows that the ingoing air is cooler than the outcoming, the cur- 
rent in the lamp' (fig. 8) is increased. If the air should require to be 
cooled, as rarely occurs, this may be effected by drawing off a portion 
of the water and replacing it by colder water. 
REFRIGERATING APPARATUS. 
The refrigerating tank is shown in fig. 10. Liquid ammonia is sup- 
plied to the tank through the iron pipe marked P5 from an ammonia 
machine in another part of the room. This liquid ammonia, initially at 
high pressure, passes through 
a reducing valve and flows into 
the coil of iron pipe shown in 
the figure. Here it evapo- 
rates, as it flows around the 
coil, as rapidly as heat can be 
absorbed from the solution of 
calcium chlorid which fills the 
tank. The temperature of this 
solution is thus reduced to 
about-18° or - 20° C. The 
coil of pipe through which the 
ammonia passes consists of 88 
meters (287 feet) of 2.5 centi- 
meters (1 inch) extra-heavy 
pipe electrically welded so as 
to form one continuous tube. 
There are two sets of air re- 
frigerators or "freezers" in 
the tank. The first is marked 
A, and is made of sheet cop- 
per. It consists practically 
of two large U tubes of cop- 
per immersed in the brine and 
so connected by horizontal elbows that the whole forms a compact mass 
a meter in height and a little over 20 centimeters square. Copper 
vanes radiating from the outside of the pipes increase the cooling sur- 
face. The air enters the pipe Pt from out-of doors laden with more or 
less moisture. Peaching the freezer A its temperature is quickly 
lowered nearly to that of the surrounding brine and nearly all of its 
moisture is deposited upon the inner walls of the freezer. Before 
the condensation of moisture is such as to threaten to close the air 
space the freezer is removed and another put in its place. The air 
leaving the freezer A passes along the pipe P2 to the calorimeter, being 
warmed to the temperature of the room before reaching the chamber. 
Its moisture content as it enters the calorimeter is quite uniform. 
Fig. 9.-"Vestibule." The object of this apparatus is to 
facilitate the comparison of the temperature of the 
incoming and outgoing ventilating currents of air. 
The air enters the lower part of the vestibule through 
the lower pipe U, and leaves the upper part through 
Uj. The thermo-electric junctions set in the partition 
between the upper and lower parts of the vestibule 
indicate any difference of temperature between the 
entering air and that which is about to flow out. 30 
Returning from the calorimeter laden with water vapor, the air passes 
through the second system of refrigerators or freezers, in which the 
greater portion of this moisture is condensed. But in order to be able 
to collect and measure the deposited moisture for any given period, 
this system of freezers is differently constructed from that used for the 
Fig. 10.-Refrigerating apparatus. The chief object of this is to remove the greater portion of the 
moisture from the ventilating current of air before it enters and after it leaves the respiration cham- 
ber. For this purpose the air is passed through copper cylinders or "coolers," A, E, immersed in 
brine (of calcium chlorid) •which is cooled to a temperature of -18° to -20° C by aid of ammonia 
expanding in the coiled pipe seen against the sides of the tank in the figure. The air flows through 
the cooler A on its way to the respiration chamber, and through B and C or D and E on its way 
from the chamber to the pumps. The valves V, and V2 are used to direct the outcoming air cur- 
rent from one pair of cylinders to the other. 
incoming air. Four cylindrical brass vessels are clamped together, as 
shown in the figure, and fitting closely inside them are four cylinders or 
freezers made of sheet copper. One of these freezers is shown in fig. 11. 
It is about 57 centimeters (22.5 inches) long and 12 centimeters (4.75 
inches) in diameter, and is divided vertically by a partition, not shown 31 
in the figure, which extends from the top nearly to the bottom. The air 
enters through a tube in the top, flows down to the bottom, and, rising 
on the other side of the partition, passes by means of a double-elbow 
connection to a second freezer, where the larger portion 
of the water not removed by the first freezer is con- 
densed. In order to give more cooling surface, vanes of 
sheet copper are soldered to the interior surface of the 
cylinders extending to within a short distance of the 
middle partition. The space between the inner cylinder 
or freezer and the outer protecting jacket is filled with 
90 per cent alcohol, in order to insure free passage of 
heat from the air current to the cooling brine. The 
freezers are removed and weighed at intervals, generally 
of six hours. In order that this can be done conveniently 
and without interfering with the regular flow of the air 
current two pairs of cylinders are used and a device is 
employed by which the current can be easily diverted 
from one pair to the other. This latter device we have 
called a "valve box." It consists, essentially, of two 
pieces of brass pipe, 10 centimeters (4 inches) in diameter, 
placed end to end. The adjoining ends are separated by 
a disk of so-called fibroid, which serves to prevent the pas- 
sage of both air and heat from one tube to the other, 
while the open ends are connected to the air pipes P3 and 
P4 (fig. 10, p. 30). The valve boxes are also provided with 
apertures and brass tubes to connect them with the 
freezers. An arrangement is made for conveniently open- 
ing and closing these apertures by rubber stoppers, which 
are held by brass rods reaching through the walls of the 
valve box and worked from the outside. Two openings on 
one side are connected with one pair of freezers, while the 
two on the other side are closed. When it is desired to 
change the freezers the current is diverted from one pair 
to the other by means of these stoppers. The first pair of freezers is 
then taken out, warmed, dried, and weighed, the increase in weight rep- 
resenting the moisture condensed from the air current during the period. 
Fig. 11.-Con- 
densing cylin- 
der of refrig- 
erating appa- 
ratus. One of 
these cylin- 
ders fits close- 
ly inside each 
of the fixed 
cylinders B, 
C, D, E, shown 
in fig. 10. The 
larger part of 
the moisture 
of theoutcom- 
ing air cur- 
rent is depos- 
ited in these 
inside cylin- 
ders. 
METER PUMP FOR REGULATING, MEASURING, AND SAMPLING THE 
VENTILATING AIR CURRENT. 
Two forms of apparatus have been used for maintaining the air cur- 
rent and measuring its volume. The first consisted of an exhausting 
air pump used in connection with a special form of gas meter made by 
Elster in Berlin. With this we were unable to make measurements as 
accurately as seemed desirable to us. 
For taking samples of air for analysis aspirators (Pl. VIII, p. 34), con- 
taining 150 liters, were employed at the outset and are still used. The 
measurements with these have been found quite accurate. The most 32 
satisfactory arrangement which we have found, however, and one which, 
serves the threefold purpose of maintaining the air current, measuring 
its volume, and delivering aliquot samples of convenient volumes for 
analyses is an apparatus designed and made by Mr. O. S. Blakeslee 
and appropriately designated by him as a "meter pump." This is 
shown in Pl. VII. The essential parts for maintaining the air current 
and measuring its volume are cylinders of steel. There are two pumps, 
which work synchronously. Three steel cylinders are employed for each 
pump. The inner and outer cylinders are arranged concentrically, with 
a narrow annular space between them. This space is nearly filled 
with mercury. Between the inner and outer cylinders, which are sta- 
tionary, plays a third cylinder, its lower portion being immersed in the 
mercury. This moving cylinder is closed at the top and is raised and 
lowered by a walking beam. The length of the stroke of the cylinder 
is determined by suitable stops, and the volume of air delivered at 
each stroke is constant. The temperature of the pump is kept prac- 
tically constant, so that the mass of air of each stroke is constant. Like 
the moving cylinder, the inner cylinder is also covered at the top, but 
through this cover are two circular apertures, opened and closed alter- 
nately by an automatic valve. 
The pump is driven by a belt on the large pulley, shown back of the 
walking beam, from a countershaft overhead The speed is reduced 
by gearing, so that the face wheel in front rotates about ten times a 
minute. On the same shaft with the walking beam is a sleeve with a 
projecting arm, which is nearly hidden behind a portion of the walking 
beam. The end of this arm is attached to the face wheel by a connect- 
ing rod, as shown in the figure. This conveys an oscillatory motion 
to the sleeve, which is transmitted to the walking beam by the straight 
steel springs, the ends of which are connected at about the middle of 
the arcs. 
The moving cylinders are suspended from the circular ends of the 
walking beam by flexible steel straps, each of which is wrapped around 
the end of the beam as the cylinder rises, as shown at the right in the 
figure. The length of the stroke is determined by stops on the rods at 
the sides of the pumps. These rods are fastened at the upper ends to 
the moving cylinders. The cylinders are thus stopped before the con- 
necting rod has reached the end of its stroke or "dead point," and 
therefore the sleeve continues to turn after the walking beam and cyl- 
inders have come to rest. This is made possible by the long steel 
springs which drive the walking beam, as they bend enough to allow 
the sleeve to turn a little after the walking beam has been stopped. 
During this short interval, while the cylinders are at rest, the valves 
are reversed by means of compressed air, so that as either cylinder 
descends the air will flow into the exit pipe. As seen in the figure, there 
are two pipes running under each pump. These pipes extend up inside 
the pump to the top of the stationary cylinders, the ends just reaching 
through the bottom of the pump proper. The valve is so arranged U. S. Dept, of Agri., Bui. 63, Office of Expt. Stations. 
Plate VII. 
Blakeslee Meter Pump. 
The dotted lines show the positions of the pipes through which the main air current enters and 
the samples for analysis pass out. 33 
that when one of these pipes is closed the other is open. In front of 
the standard carrying the driving mechanism of the pump is seen a 
small rod which extends below the base. This is driven from the 
sleeve by a pair of beveled gears, and hence it also 'has an oscillating 
movement. At the lower end it carries an arc, at each end of which 
a pin projects radially. A portion of this arc and the two pins may 
be seen near the bottom of the figure. One pin set in the slit near the 
end of the arc is crossed by the lower of the four dotted lines at the 
seventh dot from the left-hand end. The pin at the other end shows 
as a white dash a little to the right and near the second dotted line. 
Just at the right of the first pin is seen a vertical pin. In the position 
shown the pump is not quite at the end of its stroke. When the cyl- 
inders are brought to rest the moving pin will be very near this vertical 
one, and an instant later it will carry it to the right. This opens a 
port in the compressed-air chamber; the air escapes into a cylinder 
similar to one on a steam engine; the piston makes one stroke, which 
reverses the valves in both pumps, setting one so that air may be drawn 
in from the in-going pipe and the other so that the air may be expelled 
into the exit pipe. All this is done while the cylinders are at rest, and 
thus the air within them is at atmospheric pressure. The volume of 
each cylinder is known from its dimensions, and an automatic counter 
records the number of strokes. 
In front of the base of the standard are seen the electro-magnets 
used in taking the samples for analysis. Below them is the chamber 
through which all of the air from the exit pipe must pass before it 
can escape. There are three openings from this chamber-one in front 
and one at either side. From the side openings are pipes (shown by 
the dotted lines) leading to the receptacles for holding the samples 
before analysis, termed briefly "pans." Likewise a pipe placed on the 
opening in front conducts 'the main current of air to the boxes sur- 
rounding the pans. This gives the same atmosphere both within and 
without the pans, and has been found by experiment to prevent a 
transfer of moisture through the sheet-rubber tops of the pans. 
An ingenious method for taking these samples was devised by Mr. 
Blakeslee. Between the electro magnets and the standard is a ratchet 
wheel having 100 teeth. At each stroke of the pump this wheel is 
advanced by one tooth. Mounted on this wheel is a pin which closes 
an electric circuit once every revolution-that is, once in every hundred 
strokes. The current flows through one of the electro-magnets, drawing 
in its plunger. This motion closes the aperture in front, at the same 
time opening the one at the side, and thus allowing the air from a 
complete double stroke (both cylinders full) to pass into the corre- 
sponding pan. The electric circuit is then broken, which releases the 
plunger of the magnet, and the air continues to flow through the open- 
ing in front. A second pin in the toothed wheel, on the opposite side 
from the first, closes a second circuit through the other electro-magnet, 
17951-No. 63 3 34 
so that the intermediate fiftieth strokes are delivered into the other 
pan for duplicate analysis. 
THE ASPIRATORS. 
The form and use of the aspirators (Pl. VIII) were described at 
considerable length in a former publication1 and need be but briefly 
described in this connection. 
The aspirators are large cylinders made of galvanized sheet iron and 
having a volume of about 150 liters. The ends are conical, terminating 
both at the top and bottom in a small pipe. These aspirators are con- 
nected to the city water pipes, and by the opening of a valve each can 
be filled with water. When full the valve is tightly closed and the 
aspirator is ready for use. The water is drawn from the bottom of the 
aspirator by a rubber tube with a nozzle at its lower end. The flow is 
regulated by means of a pinchcock on the rubber tube and by raising 
or lowering the nozzle in order to diminish or increase the head of 
water. As the water level falls in the aspirator the nozzle must be 
lowered or the pinchcock loosened. The rate at which the water flows 
is taken every half hour, and amounts to not far from 425 cubic centi- 
meters per minute, dependent upon the exact size of the aspirator and 
the period in which it is to empty. An automatic device for maintain- 
ing a constant flow of water has been made and is now being used with 
decided success. As the water runs out air must enter the top to take 
its place. This air is drawn from one of the pipes, from which a sample 
is desired, through the absorption tubes before entering the aspirator. 
The amount of water in the aspirator at any time is shown by a glass 
gauge along the side. The volume of the aspirator is taken between a 
mark on this gauge at the top, where the cross section is very small, and 
a similar mark at the bottom. This volume has been measured many 
times and is known very exactly. 
The volume of the sample is obtained from the volume of the aspi- 
rator by adding several corrections. The temperature of the air in the 
aspirator is measured, and if it is different from that of the main cur- 
rent at the point where its volume is measured-i. e., the pump-the 
necessary correction is made. It is assumed that the air is saturated 
with water vapor, and as it entered perfectly dry a correction due to 
the pressure of the water vapor must be applied to find the volume of 
the air as it enters the aspirator. To this volume must be added the 
volume of the carbon dioxid and the water caught by the absorption 
tubes. This gives the volume of the sample taken from the main cur- 
rent, and this must be added to the volume measured by the pump to 
give the total ventilation. There are six aspirators, two being used 
for the analysis of the ingoing air and two others for the duplicate 
analysis of the outcoming air. In the meanwhile the third pair is 
being filled, so as to be ready for use when the others are empty. 
1 U. S. Dept. Agr., Office of Experiment Stations Bui. 44. U. S. Dept, of Agri., Bui. 63, Office of Expt. Stations. 
Plate VIII. 
The Aspirators. 
The aspirators are made of galvanized sheet iron and have a capacity of about 150 liters each. 
The absorbing tubes for determining carbon dioxid and water in the air drawn out of the res- 
piration chamber rest upon roof-shaped supports on the small shelves below the aspirators 
at the left. 35 
ARRANGEMENTS FOR SAMPLING AND ANALYZING THE VENTILATING 
AIR CURRENT. 
After the incoming air has left the freezer samples for analysis are 
conveyed through a copper tube, 0.5 centimeter internal diameter, to 
the apparatus for determining the carbon dioxid and water. These 
samples are drawn by two aspirators (see p. 34), each of about 150 
liters capacity. The outgoing air is sampled either by means of 
similar aspirators, or by means of the meter pump as already explained 
on page 34, or by both these methods. In the series of experiments 
carried on during the years 1896-1898 the method of sampling the out- 
going air current by means of the meter pump, and the subsequent 
analysis of this sample, had not been perfected, and the results obtained 
by the aspirators were used entirely in the calculations of the amount 
of carbon dioxid in the outgoing air and of the water vapor not 
removed by the freezers. 
The method of calculating the total carbon dioxid and water from 
the data obtained by use of the aspirators is as follows: 
Let v = volume of aspirator in liters. 
v' = corrected volume of aspirator in liters. 
B = observed barometer. 
b = observed barometer corrected for tension of aqueous vapor 
in aspirator. 
T = temperature of pump, 
t = temperature of aspirator. 
w = total weight, 'in milligrams, of carbon dioxid or water 
removed from air in aspirator. 
vw - equivalent volume of water vapor in liters (weight in grams 
of water absorbed multiplied by 1.33). 
vc = equivalent volume of carbon dioxid in liters (weight in 
grams of carbon dioxid absorbed multiplied by 0.566). 
nq = milligrams of carbon dioxid or water per liter. 
Vp = volume, in liters, of air passed through pump. 
V = total ventilation in liters = V 4- n1 (where analysis is in 
duplicate, vz here represents the sum of the volumes of 
the aspirators used). 
W = total weight of carbon dioxid or water in ventilating air 
current. 
1 > • 
Then: vz = v. [1 + a (T - t)] 4- vw 4- vc. • 
w 
ml=-. 
V = Vp +vz. 
W = V mp 
The method for collecting the sample delivered by the meter pump 
is as follows: The air from each fiftieth stroke of the pumps is col- 36 
lected alternately in the "upper" and "lower" pans. These pans, 
which are made of heavy sheet tin, are 41 centimeters (16 inches) in 
diameter at the top, 28 centimeters (11 inches) at the bottom, and are 
13 centimeters (5 inches) deep. A light, flexible rubber diaphragm 
made of the same dimensions exactly fits the interior of the pan, and 
its upper edge is securely fastened to the rim of the latter by means of 
a broad, stout rubber band. The air enters between the pan and the 
rubber diaphragm, and as the sample is delivered into the pan the 
diaphragm rises. The maximum capacity of the pan is such that it is 
possible for two samples to be delivered before any air is removed 
from it. At this point the rubber diaphragm stands inverted over the 
pan, and the whole presents the apperance of a double pan. In actual 
use the pans are not allowed to fill to the maximum, nor are they 
entirely emptied except at the close of the six-hour experimental 
periods. 
The air is drawn from the pans by suction created by a small air 
pump. This pump partially exhausts the air in a large bottle, which 
serves as a regulator, so that the suction will be more constant. From 
this regulator rubber tubes lead to two Wolff bottles, one for each pan. 
These Wolff bottles are partly filled with water. Dipping below the 
surface of the water are two glass tubes, one of which opens directly 
into the air, while the other is connected with the tubes used for the 
absorption of the carbon dioxid and water vapor of the air coming 
from the pans. The tube opening directly into the air can be raised 
or lowered until the air is drawn from the pans at the desired rate- 
that is, sufficiently fast nearly to empty them before the delivery of 
the next sample. The air leaves the pans through a stopcock at the 
bottom and passes through the apparatus for the removal of the carbon 
dioxid and water. 
If the pans were allowed to become entirely empty, and at the same 
time the suction were kept up, harm might result. To obviate this 
danger an automatic regulator was devised which completely shuts off 
the air passage from the Wolff bottles to the pans at a point immedi- 
ately beyond the absorption tubes. Inasmuch as the pans were not 
used in the experiments of 1896-1898, with two exceptions (experiments 
Nos. 12 and 13), the self-regulating safety device is not described in 
this publication. To avoid danger of the air being withdrawn from the 
pan so slowly that a second sample maybe delivered before the preced- 
ing one is removed, the rate of the current of air is so regulated by 
means of the Wolff bottles that the air will be withdrawn to the desired 
point and the automatic device will shut off the current several minutes 
before the time for the delivery of the next sample 
The method of analysis by use of the pans need not be further 
described in this publication. 
The method of analysis of the air current consists in passing the 
sample through u tubes containing sulphuric acid for the removal of 
the water and soda-lime for the removal of the carbon dioxid. These 37 
U tubes are about 12 centimeters (5 inches) in length and 1.5 centime- 
ters (| inch) in diameter. Those for the absorption of water are about 
two thirds filled with pumice stone drenched with concentrated sul- 
phuric acid. In use these tubes are laid nearly flat, with just enough 
incline to the tube to avoid all danger of the acid coming in contact 
with the rubber stoppers through which the small connecting glass 
tubes pass. Detailed experiments have shown that the absorption of 
water vapor from an air current flowing at a rate as high as 750 cubic 
centimeters per minute is complete by the use of one such sulphuric- 
acid tube until the acid has taken up 1 gram of water, after which 
complete absorption from a current flowing at this rate can not be 
depended upon. 
While only one □ tube is necessary for the complete absorption of 
the water, a system of three □ tubes is used for the complete removal 
of the carbon dioxid from the sample. Two of these are filled with 
soda-lime for the absorption of the carbon dioxid itself; the third con- 
tains sulphuric acid, which serves for the collection of the water taken 
up from the soda lime by the current of perfectly dry air passing over 
it. The soda-lime is made in the manner described in a former publi- 
cation.1 As the absorption of carbon dioxid takes place the yellowish 
color of the soda-lime gives place to the white color of the sodium and 
calcium carbonates that are formed, thus making it easy to see when a 
fresh tube is necessary. In the analysis of the incoming air, which 
contains very small quantities of carbon dioxid, one soda-lime tube, 
followed by a sulphuric-acid tube, is found sufficient, but for the out- 
going air three tubes are used, as above described. When the soda- 
lime in one is nearly saturated the tube is removed, the second moved 
forward to take its place, and a fresh tube is added in place of the 
second. In this way the second tube in the system is always filled 
with fresh soda-lime, ft has been found that two soda lime tubes 
will remove all of the carbon dioxid from an air current flowing at 
a rate as high as 750 cubic centimeters per minute. 
RESIDUAL CARBON DIOXID AND WATER VAPOR IN THE RES- 
PIRATION CHAMBER. 
The term "residual carbon dioxid and water vapor" is applied to 
the total amount of these materials remaining in the chamber at any 
given time. 
In the check experiments described beyond, in which alcohol was 
burned, and in the experiments with man, the quantities of carbon 
dioxid and water vapor in the air of the chamber were found to be 
subject to considerable variations. In the experiments in which alco- 
hol was burned in the chamber these fluctuations are due to the com- 
bustion of the alcohol at different rates; in experiments with a man 
they are due to differences in the amount of external or internal 
1 U. S. Dept. Agr., Office of Experiment Stations Bui. 44. 38 
work done, and the consequent changes in the amount of material 
oxidized and the carbon dioxid and water eliminated. 
In order to determine the exact amounts of carbon dioxid and water 
given off during any given time from the combustion of alcohol in a lamp 
or of food in the body, it is necessary to know not only the amount of these 
respiration products in the ventilating air current during this period, 
but the amount in the chamber at the beginning and at the end. For 
this purpose a sample consisting of 10 or 12 liters of air is drawn from 
the interior of the chamber by means of a rubber tube passing through 
an opening in the cover of the food aperture. The number of milli- 
grams of carbon dioxid and water vapor per liter found in the sample, 
multiplied by the total amount of air in the chamber (approximately 
4,800 liters), gives the total amounts of these products in the chamber 
at the time the sample was drawn. Such analyses are made at the 
beginning of an experiment and at the close of each experimental 
period. 
An increase in the residual carbon dioxid and water vapor during 
any experimental period indicates that more of these products were 
given off in this time than during the previous period, and that some 
of the excess has remained in the chamber. Conversely, a decrease in 
the carbon dioxid and water vapor during any experimental period 
indicates a smaller production of these products and the consequent 
removal of some of the substances in the chamber at the beginning of 
the period. In other words, the analysis of the air in the chamber 
gives data for determining a positive or negative correction which 
must be applied to the amount of respiratory products found in the 
ventilating air current in order to obtain the quantities of these prod- 
ucts actually given off during any given time. 
TESTS OF THE ACCURACY OF THE APPARATUS AS A CALORIMETER. 
Two series of test .experiments were instituted to ascertain whether 
the calorimeter would accurately measure a given quantity of heat gen- 
erated within the respiration chamber. In the first series an electric 
current of measured strength flowing through a known resistance 
generated in a given time a quantity of heat which could be readily 
calculated. As there was no circulation of air and no water vapor 
produced, the problem was simplified. In the second series alcohol 
was burned within the chamber, a current of air circulating through it 
supplying the necessary oxygen and carrying away the products of 
combustion. Here the complications arising from a ventilating current 
of air and the production of a large amount of moisture were introduced. 
Evidently, if the calorimeter will accurately measure a known quantity 
of heat generated in these two ways, it may be employed with confidence 
to measure the treat generated by a living subject within the chamber. 
In the latter case, however, we can not expect as high a degree of 
accuracy as in the former, for the rate of evolution of heat is more vari- 39 
able and the movement of the person or animal about the respiration 
chamber introduces disturbances which will prevent the highest degree 
of accuracy in the measurements; and furthermore the errors due to 
physiological causes may be considerable. 
ELECTRICAL TEST EXPERIMENTS. 
The arrangement of the apparatus in the electrical tests is shown in 
fig. 12. The current from the 220-volt mains passes into the respiration 
chamber through a variable resistance, RS, a Weston ammeter, Am, 
and two copper voltameters, Vm, where it passes through PQ, a coil 
of German-silver wire No. 30, having a resistance of about 100 ohms. 
A Kelvin balance in series with a known resistance, Rb in oil, and a 
Kelvin Balance 
Fig. 12.-Arrangement of electric apparatus for generating heat in the test experiments. 
Weston voltmeter in parallel with them are connected to the points 
a, b, and indicate the fall of potential through the resistance PQ 
Usually an assistant keeps the current constant by adjusting the 
resistance RS. The current is given with considerable accuracy by 
the Weston ammeter, which in the first test experiment was used 
alone. But when the copper voltameters were employed the values for 
the current obtained with them were used. 
The original data for the calorimetric measurements as observed by 
the operator at the observer's table are recorded by him in a notebook 
especially prepared for the purpose. 
A sample page from the record of the test of March 20,1897, is given 40 
below. It gives the observations of one hour. The Weston instru- 
ments were used to measure the current and voltage, and the current 
was not maintained constant. Readings were recorded every two 
minutes, and the averages taken for the values of the current and the 
voltage. 
Table 1.-Sample page from record of calorimetric observations of experiment of 
March 20, 1897. 
Time 
p. m. 
Inner wall. 
Water circuit. 
Inside 
temper- 
ature 
No. 5. 
Electrical 
measure- 
ment. 
Remarks. 
No.l. 
Parts. 
T,-T2. 
Tj-Tp 
Bridge 
No. 6. 
h. 
m. 
°C. 
°C. 
°C. 
°C. 
Volts. 
Amp. 
-11 
( 0 
3 
00 
0 
+4 
0 
9.62 
19.95 
71 
0. 61 
02 
i 
0 
9.60 
19.95 
71 
.61 
04 
i 
9.60 
19.95 
71 
.61 
06 
1 
13. 58 
9.61 
19. 95 
71 
.61 
08 
i 
4.14 
9. 44 
9.60 
19. 95 
70 
.61 
10 
0 
9.57 
19.95 
71 
.61 
12 
i 
9.57 
19.95 
72 
.61 
14 
13.54 
9.54 
19. 95 
72 
.61 
5th 10 kilograms at 3.14. 
16 
i 
4.14 
9.40 
9. 52 
19.95 
71 
.61 
Average bridge reading, 9. 56. 
18 
i 
9.50 
19.95 
72 
.61 
9.56 - 0.13=9.43°. 
20 
0 
0 
9. 54 
19. 95 
72 
.61 
9.43° X 10 = 94.30 calories. 
22 
i 
13.57 
9.54 
19.95 
70 
.61 
24 
1 
4.15 
9.42 
9. 51 
19.94 
69 
.60 
26 
1 
9.56 
19.94 
70 
.59 
28 
9. 57 
19.93 
69 
.59 
30 
2 
fill 
13.58 
9.58 
19.92 
70 
.60 
32 
। ° । 
4.14 
9.44 
9.51 
19.92 
69 
.61 
34 
i 
l-i J 
9.53 
19. 92 
71 
.61 
36 
0 
9.51 
19. 92 
71 
.61 
38 
0 
13.57 
9.53 
19.92 
71 
.61 
40 
i 
4.14 
9.43 
9.50 
19.92 
71 
.61 
42 
i 
9. 54 
19.93 
71 
.61 
44 
i 
( 0 
9.58 
19 93 
71 
.61 
46 
0 
+1J 
- 4 
13.57 
9.55 
19.94 
71 
.61 
48 
i 
[-24 
4.16 
9.41 
9.55 
19.94 
74 
.63 
50 
0 
9.46 
19.95 
73 
.63 
52 
0 
9.46 
19.96 
73 
.63 
54 
J 
13.54 
9.46 
19.96 
74 
.63 
56 
i 
4.17 
9. 37 
9.49 
19.97 
73 
.62 
58 
J 
9. 39 
19.97 
.62 
In the first column the time is recorded. In the second column are 
the deflections, in scale divisions, of the galvanometer when joined to 
thermo-electric circuit (No. 1) of the walls of the respiration chamber. 
Numbers at the left side of the column indicate deflections to the left 
and show that the inner chamber is a little cooler than air space B. 
(See Pls. IV and V.) At 3.10, the air in B having been cooled slightly, 
the deflection is zero. Then it becomes positive for a few minutes and 
the air in B requires slight warming. The deflections during the hour 41 
are on the average less than one division and indicate an average dif- 
erence of temperature of about 0.004° C. But this difference is part of 
the time positive and part of the time negative, so that the algebraic 
average is only one-fifth of one division to the right, indicating that 
the inner chamber was on the average for the hour 0.0014° C. warmer 
than the surrounding space. By carrying these differences along from 
page to page, the adjustments are made so that the differences cancel 
each other, and there is no correction to apply for flow of heat through 
the walls for the whole experiment. 
One division deflection corresponds, as already stated (p. 18), to 0.007° 
C. difference of temperature, and with this difference of temperature 4 
small calories per minute, or 0.24 of a large calorie per hour, flow 
through the walls of the chamber. This was determined in the follow- 
ing manner: A current of 0.17 amperes was passed through a resistance 
within the calorimeter for four hours. The fall of potential was 43 volts, 
giving 7.3 watts or about 104 small calories per minute. The tempera- 
ture of the chamber rose a little at first, but was constant during the 
latter part of the experiment, showing that all of the heat was escaping 
through the walls. The deflection produced by the thermo-electric ele- 
ments of circuit No. 1 was about 25 divisions, remaining substantially 
constant for sometime before the end of the experiment. This gives 
4 small calories per minute per scale division as the amount of heat 
flowing through the walls of the calorimeter. This may be called the 
radiation constant. 
For an average deflection of one fifth of one division the rate at which 
heat would be escaping from the interior of the calorimeter would be 
0.048 calorie per hour. The amount of heat generated and carried 
away from the calorimeter was, in this experiment, 74 calories per 
hour. Thus the amount of heat passing through the walls is only 
0.048 4- 74 = 0.00065 of the total, or one sixteenth of 1 per cent, and 
would not be an appreciable error if uncorrected. It is, however, 
subsequently recovered by keeping the deflection slightly negative, as 
already explained. 
In the third column are recorded the deflections observed when parts 
of the thermo-electric circuits are separately joined to the galvanometer, 
the four numbers being for the top, upper, lower, and bottom sections, 
respectively. These deflections are all small and make more certain 
the balance indicated by zero deflection of the entire system in series. 
In the fourth column are recorded the readings of the two mercury 
thermometers, G and H (fig. 3), and the fifth column shows their differ- 
ence (T2-Ti), which is the gain in temperature of the water flowing 
through the absorbing pipes. The sixth column shows the readings of 
bridge No. 6 for the same difference of temperature. 
The bridge does not give the exact difference of temperature of the two 
coils, which are immersed in the water, its readings being always too high. 
However, it correctly indicates the variations of temperature between 42 
the thermometer readings, and thus interpolations between the observed 
thermometer readings can be made. In actual practice the average 
of the bridge readings is reduced by the average difference between 
thermometer and bridge. For example, 10 kilograms ran out during 
the period of seventy-four minutes from 2 to 3.14 p. m. The average 
of the bridge readings for the seventy-four minutes, during which 
the 10 kilograms of water were passing through the absorbers, was 
9.56. The correction to be subtracted to reduce the bridge readings 
to degrees centigrade (see p. 40) is 0.13. This gives an average differ- 
ence of 9.43°, and hence 94.30 calories of heat were brought away in 
these seventy-four minutes. In the same way the difference of temper- 
ature for each 10 kilograms is determined. A tabular statement of the 
heat measured during the entire experiment from 9.07 a. m. to 10.26 
p. m. is given in the following table: 
Table 2.-Summary of heat measurements, March 20, 1897. 
Lime. 
(a) 
Dura- 
tion 
of 
period. 
Calorimetric measurements. 
Electrical measurements. 
Tempera- 
ture of 
the inside 
by the 
copper 
thermom- 
eter No. 5. 
(b) 
Water. 
(c) 
Differ- 
ence in 
tempera 
ture by 
bridge 
No. 6. 
(d) 
Cor 
rected 
difference 
in 
tempera- 
ture, 
c-0.13. 
(e) 
Amount 
of 
beat, 
b x d. 
E 
Volt- 
age on 
coil. 
I 
Cur- 
rent 
through 
coil. 
W 
Watts, 
E X I 
(k) 
Calories, 
W X a x 
.2378. 
h. 
m. 
s. 
Secs. 
Kilos. 
°C. 
° C. 
Cals. 
Volts. 
Amp. 
°C. 
9 
07 
00 
20. 10 
9 
51 
15 
2, 655 
71 
9.57 
9.44 
66. 08 
148.3 
0.625 
92.69 
58. 52 
19. 78 
10 
12 
30 
1. 275 
31 
9.07 
8.94 
26.82 
146.4 
.618 
90. 48 
27. 43 
19. 82 
11 
11 
00 
3,510 
8( 
9.51 
9.38 
75.04 
142.7 
.604 
86.19 
71.94 
19.76 
11 
28 
00 
1,020 
21 
9.76 
9. 63 
19. 26 
141.5 
.596 
84.33 
20.45 
19. 84 
12 
46 
00 
4.680 
10 
9. 76 
9.63 
96.30 
141.7 
.602 
85.30 
94.93 
19 93 
2 
00 
00 
4,440 
10 
9.58 
9.45 
94.50 
145.2 
.617 
89.59 
94.59 
19 93 
3 
14 
00 
4,440 
10 
9. 56 
9.43 
94.30 
144.7 
.617 
89.27 
94.25 
19.95 
4 
28 
15 
4,455 
10 
9. 46 
9.33 
93. 30 
143.1 
.604 
86. 43 
91.56 
19.91 
5 
41 
45 
4,410 
10 
9. 37 
9. 24 
92. 40 
143.9 
.612 
88.07 
92. 36 
19.92 
6 
56 
00 
4, 455 
10 
9.41 
9.28 
92.80 
143.9 
.612 
88.07 
93.30 
19. 94 
8 
10 
15 
4,455 
10 
9. 21 
9.08 
90.80 
142.3 
.606 
86.23 
91.35 
20.01 
9 
24 
15 
4, 440 
10 
8.98 
8.85 
88. 50 
139.5 
.594 
82. 86 
87.49 
19.99 
10 
26 
00 
3, 705 
8 
9 10 
8.97 
71.76 
1.001.86 
138.0 .583 80.45 
Heat generated by elec- 
70. 88 
19. 95 
Capacity correction for 0.15°.... 
Heat measured by calorimeter.. 
= -9. 00 
= 992. 86 
trie current. 
= 989. 05 
When the rate of flow of the water is changed during the time of 
filling one water meter (10 kilograms), the difference of temperature is 
averaged for each part separately. Thus, in the first period of this 
experiment, at 9:51:15 (see table above), when 7 kilograms had been 
collected the rate was decreased. Hence the difference of temperature 
for the 7 kilograms and for the 3 kilograms are separately determined, 
and the sum of the corresponding amounts of heat (66.08 and 26.82) is 
the total for the 10 kilograms. 43 
Column 7, Table 1, shows the inside temperature as determined by 
bridge No. 5. The eighth column shows the readings of the voltmeter. 
There was a resistance of 20,000 ohms in series with the voltmeter, the 
resistance of which was 19,210; hence its readings should be multiplied 
by 392104-19210=2.041. The average of the voltmeter readings for 
the 10 kilograms from 2 p. m. to 3.14 p. m. was 70.9; multiplying this 
by 2.041 gives 144.7, which appears in the seventh line of the seventh 
column of Table 2, page 42. The average current for the same period 
was 0.617 ampere, and is entered in the eighth column, page 42. The 
ninth column gives the product of current and electromotive force, or 
watts; and the tenth column the number of calories of heat, found by 
multiplying watts by the time in seconds and the constant factor 
0.2378. 
If we denote by C the current in amperes, by E the electromotive 
force in volts, and by t the time in seconds, then 
Ox 10-1=the current in absolute C. G. S. units. 
and E x 108=the electromotive force in C. G. S. units. 
The rate of work or the electrical power is given by the product of 
the current and the electromotive force, 
= CE in watts, 
=CE x 107 in C. G. S. units or ergs per second. 
Hence in the time t, 
energy in ergs=CEtx 107. 
If a calorie be defined as the amount of heat required to raise 1 
kilogram of water 1 degree at 15° C. it will be equivalent to a definite 
number of ergs, which must be determined by experiment. The value 
given by E. H. Griffiths,1 as the result of his elaborate work on the 
determination of the relation between the electrical units and those of 
beat is, 
J = 4.1982 x 107 ergs, at 15° C. 
The average temperature of the water in the absorbers in our experi- 
ments was in every case between 8° and 10° C. The specific heat of 
water2 at 10° C. is 1.00193 times its value at 15° C, and hence it 
requires 1.0019 times as much heat to raise 1 kilogram 1 degree at 10° 
C. as at 15° C., so that 
J = 4.2062 x 107 ergs, at 10° C. 
This value is practically the same as that given by Professor Row- 
land4 for the mechanical equivalent of heat at 8° C., and so appears to 
be the most nearly correct for our use. 
1 Phil. Trans. Royal Society [London], A, 1893. Proc. Roy. Soc., vol. 55. 
2 Bartoli and Stracciati, Bol. Men. Accad. Gioenia, 18, 1891, Apr. 26. Preston's 
Theory of Heat, p. 265. 
3 As estimated from the figures of the table on p. 56-i. e., 1.0029 = 1.0010 = 1.0019. 
4 Proc. Amer. Acad. Sci., 15 (1879), p. 75. 44 
A recent recalculation of Rowland's value of the mechanical equiva- 
lent of heat1 gives 4.200 x 107, at 8° C., and it has been suggested that 
the difference between this value and Griffith's is accounted for by the 
uncertainty in the value of the ampere, but as we use the same ampere 
as Griffith did we should use his value of J. 
The energy of the electric current expressed in ergs being 
CEt x 107, 
when expressed in calories becomes 
(j x 0 23775 
4.2062 x 107 ~ 1 x • ' • 
Referring to Table 2, page 42, it will be seen that the total amount of 
heat generated by the current is 089.05 calories. The amount measured 
is 1,001.86 calories, exceeding that produced by the electric current by 
12.8 calories. During the experiment the temperature of the chamber 
fell from 20.10° to 19.95°. The capacity for heat of the apparatus is 
equivalent to that of 60 kilograms of water-that is, it requires 60 
calories to raise the temperature 1°. This was determined in the 
following manner: 
The calorimeter was held at a constant temperature (23.8°) for sev- 
eral hours. A current of 1.77 amperes, at 33 volts, was then passed 
through it for two hours. The temperature slowly rose and, keeping 
the deflection of the No. 1 thermo electric circuit zero, no heat was 
allowed to pass through the walls. At the end of this time the cur- 
rent was stopped and the calorimeter came to a constant temperature 
at 25.4°. The rise in temperature was thus 1.6°. The amount of heat 
generated by the electric current was 98.2 calories, or 61.5 calories per 
degree. Other determinations gave similar results, and the round 
number 60 was taken as a sufficiently exact value of the thermal 
capacity of the apparatus. 
In the above case, where the temperature fell 0.15°, there will be 
0.15° x 60 = 9.0 calories which are absorbed and measured in addition 
to the heat from the current. Applying this correction, we have for the 
measured amount of heat produced by the current, 1,001.86 - 9.0 = 
992.86 calories, which is 3.4 calories, or 0.34 per cent, greater than indi- 
cated by the electrical determination. As the temperature was falling 
more rapidly at the beginning than at the end, the actual temperature 
of the apparatus was somewhat higher than that of the air inside, and 
so the actual fall in temperature might be slightly greater than 0.15°. 
That the temperature of the air does not truly represent that of 
the apparatus when the latter varies rapidly is clearly shown at the 
beginning of this experiment. From 9.07 a m. to 9.51 a. m. the tempera- 
ture of the air fell 0.32°, which would correspond to 20 calories; but the 
extra amount of heat for that period is only 66.1 - 58.4 = 7.7 calories, 
which indicates that the temperature of the calorimeter as a whole 
fell only 0.13°. 
1 Phil. Mag., 44 (1897), p. 169. 45 
Taking the latter part of the experiment, from 12.46 p. m. to 16.26 p. m., 
when the temperature was nearly constant, the heat from the current 
is 716.3, while that measured is 718.4. As the temperature was 0.02° 
higher at the end than at the beginning, 1.2 calories must be added to 
this, making it 719.6, which is 3.3 calories, or 0.46 per cent, too great. 
However, as only the Weston instruments were used to measure the 
electrical energy in this experiment, this is as good an agreement as 
could be expected. 
In the next experiment, which was made on March 25, the current 
was kept constant, and the voltage measured by a Kelvin balance in 
series with a resistance.. An assistant kept the balance at zero by 
varying the resistance RS, fig. 12, in series with the coil inside. The 
Weston instruments were also in the circuit. The current was 
determined by a copper voltameter. 
A sample page of the records of this experiment is shown in Table 3: 
Table 3.-Sample page from record of calorimetric observations of experiment of 
March 25, 1897. 
Time 
p. m. 
A. m. 
4 00 
02 
04 
06 
08 
10 
12 
14 
16 
18 
20 
22 
24 
26 
28 
30 
32 
34 
36 
38 
40 
42 
44 
46 
48 
50 
52 
54 
56 
58 
Inner wall. 
Water circuit. 
Inside 
tempera- 
ture 
No. 5. 
°C. 
21.33 
21.33 
21.33 
21.32 
21. 32 
21. 32 
21.32 
21.32 
21.32 
21. 31 
21.31 
21.31 
21. 31 
21. 31 
21.30 
21. 29 
21.29 
21.29 
21. 29 
21.29 
21.31 
21.31 
24.31 
21.31 
21.31 
21.31 
21.31 
21.32 
21.32 
21.32 
Remarks. 
5 kilograms at 4:06:45 at 9. 68. 
Average bridge reading, 9.68. 
9.68 - 0.16 = 9.52°. 
9.52° X5=47.6 calories. 
5 kilograms at 4:38:15 at 9.55. 
Average bridge reading, 9.55. 
9.55 - 0.16 = 9.39°. 
9.39° X5 = 46.95 calories. 
No. 1. 
-21J 
4 
0 
i 
0 
0 
4 
1 
li 
0 
4 
4 
0 
0 
4 
4 
4 
4 
4 
4 
0 
0 
i 
i 
i 
2 
32J+74 
Parts. 
T^Tj. 
°C. 
T^Tp 
°C. 
Bridge 
No. 6. 
°C. 
9.65 
9.64 
9. 60 
9.60 
9.57 
9.57 
9. 55 
9. 50 
9. 47 
9.44 
9. 38 
9.42 
9.44 
9.64 
9.49 
9.43 
9. 55 
9.68 
9. 75 
9.87 
9.81 
9.88 
9. 88 
9.88 
9.82 
9.88 
9. 86 
9. 95 
10.01 
10.02 
13.12 
3. 76 
9. 36 
f -1 1 
1 0 ( 
1 +1* f 
( -3 J 
13. 22 
3.85 
9.37 
13.52 
3.75 
9. 77 46 
The amount of heat passing through the walls is very small, as 
shown by the deflections of the thermo-electric circuit No. 1. The 
heat was brought away and measured at very nearly the rate at which 
it was produced, as is evident from the constancy of the temperature 
as shown by No. 5. The increase in the bridge readings toward the 
end of the hour is due to the fact that the water was flowing more 
slowly. 
The summarized results of this experiment were as follows: 
Time. 
(a) 
Dura- 
tion of 
period. 
Calorimetric measurements. 
Electrical measurements. 
Tem- 
pera- 
ture 
of the 
inside 
by the 
copper 
ther- 
mome- 
ter, 
No. 5. 
(b) 
Water. 
(c) 
Differ- 
ence in 
tempera- 
ture by 
bridge 
No. 6. 
(d) 
Corrected 
differ- 
ence in 
tempera- 
ture, 
c -0.16. 
(e) 
Amount 
of 
heat, 
bxd. 
E 
Voltage 
on 
coil. 
I 
Current 
through 
coil. 
w 
Watts, 
Ex I. 
(k) 
Calories, 
WX a 
X .2378. 
h. 
11 
m. 8. 
17 00 
Sect. 
Kilos. 
°C. 
°C. 
Cals. 
Amp. 
°C. 
21. 42 
11 
42 30 
1,530 
41 
9.63 
9.47 
37.88 
100.58 
0.996 
100.18 
36. 45 
21.43 
12 
08 30 
1,560 
4(1 
9.56 
9.46 
37.60 
100.58 
.996 
100.18 
37.16 
21. 44 
12 
22 00 
810 
2J 
9.63 
9.47 
18.94 
100. 58 
.996 
100.18 
19.29 
21.44 
1 
13 00 
3,060 
% 
9. 49 
9. 33 
74.64 
100. 58 
.996 
100.18 
72.90 
21. 38 
1 
26 00 
780 
2/ 
9.47 
9.31 
18.62 
100.58 
.996 
100.18 
18.58 
21.38 
1 
50 00 
1,440 
41 
9. 39 
9.23 
36.92 
100. 58 
.996 
100.18 
34.31 
21. 40 
2 
16 00 
1,560 
4KJ 
9. 46 
9. 30 
37.20 
100. 58 
.996 
100.18 
37.16 
21.43 
2 
29 45 
825 
2J 
9.52 
9. 36 
18. 72 
100. 58 
.996 
100.18 
19.64 
21.41 
3 
02 15 
1,950 
5L 
9.53 
9. 37 
46.85 
100. 58 
.996 
100.18 
46.46 
21.46 
3 
34 30 
1.935 
5r 
9. 71 
9.55 
47.75 
100.58 
.996 
100.18 
46.10 
21.33 
4 
06 45 
1,935 
9.68 
9.52 
47. 60 
100.58 
.996 
100.18 
46.10 
21.32 
4 
38 15 
1,890 
5/5 
9.55 
9.39 
46.95 
100. 58 
.996 
100.18 
45. 03 
21.29 
5 
22 15 
Capa 
Heat 
2,640 
city corr 
measure 
6^6 
ection f< 
d by cal 
10.09 
>r 0.10° ... 
orimeter . 
9.93 
59.58 
529. 25 
- -6. 00 
= 523. 25 
100.58 
Heat 
elect 
.996 
jenera 
4c curn 
100.18 
ted by 
mt 
62. 89 
= 522.07 
21.32 
Table 4.-Summary of heat measurements, March 25, 1897. 
The electrical energy was much more carefully determined than 
before, and was maintained constant throughout the experiment. The 
heat as measured and corrected is practically identical with that calcu- 
lated from the current. 47 
Another experiment was made the next day, March 26, in which a 
larger amount of heat was generated and measured. This experiment 
began in the morning and extended till late in the evening. During 
the first part, however, there was some uncertainty in the measure 
ment of the temperature: accordingly, only the last seven nours are 
given. A sample page of the record and a summary of results are 
given in Tables 5 and 6. 
Table 5.-Sample page from record of calorimetric observations of experiment of 
March 26, 1897. 
Time 
p. m. 
Inner wall. 
Water circuit. 
Inside 
temper- 
ature 
No. 5. 
Remarks. 
No. 1. 
Parts. 
T,-T2. 
T2-T,. 
Bridge 
No. 6. 
h. 
m. 
°C. 
°C. 
°C. 
°C. 
-3i 
( +2 1 
5 
00 
i 
J +1 I 
1 0 f 
12.80 
22.61 
02 
0 
( -14 J 
12. 80 
22.61 
04 
0 
12. 79 
22. 61 
1st 10 kilograms at 5. 04. 
06 
0 
12. 80 
22.62 
Average bridge reading, 12 73. 
08 
1 
12.87 
22.62 
12.73 - 0.27 = 12.46°. 
10 
0 
12. 85 
22.62 
12.46° X 10 = 124.6 calories. 
12 
li 
15.97 
12.84 
22. 61 
14 
li 
3. 42 
12.55 
12.78 
22.62 
16 
'4 
12.72 
22.62 
18 
li 
12.80i 
22. 62 
20 
1 
(+1 1 
12.81 
22.62 
22 
i 
J +14 I 
1 -1 ( 
12. 84| 
22. 62 
24 
0 
I o J 
12. 83J 
22. 62 
26 
0 
12.86 
22.62 
• ■ 
28 
0 
- - 
15.93 
12. 76 
22. 63 
30 
4 
3.44 
12.49 
12. 77 
22.63 
32 
0 
12. 871 
22. 63 
34 
4 
12. 77 
22. 64 
36 
1 
12 80£ 
22. 64 
38 
0 
12. 81 
22. 64 
40 
1 
12.81 
22. 64 
42 
4 
15.94 
12.81 
22.64 
44 
4 
( 0 1 
3.43 
12.51 
12. 77 
22.64 
46 
0 
J -i I 
12 78 
22. 65 
2d 10 kilograms at 5:47:45. 
48 
1 
l -4 J 
12.79 
22.65 
Average bridge reading, 12.81. 
50 
4 
12.86 
22.65 
12.81-0.27 = 12.54°. 
52 
0 
12.83 
22.65 
12.54° X 10 = 125.4 calories. 
54 
i 
12. 79 
22. 65 
56 
0 
16.04 
12.81 
22. 64 
58 
0 
3. 47 
12.57 
12. 85 
22.64 
■ - 
7+114 48 
Table 6.-Summary of heat measurements, March 26, 1897. 
Time. 
(a) 
Dura- 
tion 
of 
period. 
Calorimetric measurements. 
Electrical measurements. 
Tempera- 
ture of 
the 
inside 
by the 
copper 
thermom- 
eter 
No. 5. 
(b) 
Water. 
(c) 
Differ- 
ence in 
temper 
ature 
by 
bridge 
No. 6. 
(d) 
Cor- 
rected 
differ- 
ence in 
temper- 
ature, 
c -0.27. 
(e) 
Amount 
of 
heat, 
bxd. 
E 
Voltage 
on 
coil. 
I 
Cur- 
rent 
through 
coil. 
W 
Watts, 
E x I. 
(k) 
Calories, 
W x a 
X.2378. 
h. 
4 
m. s. 
20 20 
Secs. 
Kilos. 
°C. 
°C. 
Cals. 
Amps. 1 
°C. 
22. 63 
5 
04 00 
2, 620 
10 
12.73 
12.46 
124.60 
147.8 
1.363 
201.5 
125.54 
22. 61 
5 
47 45 
2, 625 
10 
12.81 
12. 54 
125. 40 
147.8 
1.363 
201.5 
125. 78 
22.65 
6 
31 45 
2, 640 
10 
12.92 
12.65 
126. 50 
147.8 
1.363 
201.5 
126. 50 
22.67 
7 
15 27 
2,622 
10 
12. 77 
12. 50 
125. 00 
146 0 
1.350 
197.1 
122.89 
22. 57 
7 
59 10 
2,623 
10 
12. 80 
12.53 
125. 30 
147.8 
1.363 
201.5 
125. 68 
22. 59 
8 
42 40 
2,610 
10 
12. 79 
12.52 
125.20 
147.8 
1.363 
201.5 
125.06 
22.60 
9 
26 15 
2,615 
10 
12. 78 
12.51 
125.10 
147.8 
1.363 
201.5 
125. 30 
22.60 
10 
09 50 
2, 615 
10 
12. 68 
12.41 
124.10 
147.8 
1.363 
201.5 
125. 30 
22. 64 
10 
53 30 
2, 620 
10 
12.82 
12.55 
125. 50 
147.8 
1.363 
201.5 
125.54 
22. 62 
11 
37 10 
Capa 
Heat 
2,620 
city corr 
measure 
10 1 12.81 
ection for 0.02°. 
d by calorimeter 
12. 54 
125. 40 
1, 252.10 
= -1.2 
=1,250. 9 
147.8 1 1. 363 201. 5 
Heat generated by elec- 
tric current = 
125.54 
=1,253.13 
22.61 
The current was maintained constant, as in the preceding experi- 
ment, and the voltage measured by the Kelvin balance. About 7 p. m. 
the voltage on the mains was lower than usual for some minutes, and 
the variable portion of the resistance was insufficient to maintain the 
voltage on the coil. Consequently the balance was set at a lower 
point and the current kept constant at this smaller value until it could 
be restored to its former value. This makes the average voltage and 
current for the fourth period smaller than the others. The totals show 
that the amount measured was 2.2 calories, or 0.18 per cent, too small. 
This, however, assumes that the measurements of temperature at the 
beginning and the end are accurate and that the temperature of the 
whole apparatus is the same as that of the air. An error of a few 
hundredths of a degree might throw the difference in the opposite 
direction. Since this change is so slight, and as all the measurements 
of the temperature are presumably equally reliable, it would be more 
accurate if the average of all was used in obtaining the capacity cor- 
rection. This can be done by taking different portions of the experi- 
ment and applying the capacity correction to each portion; the average 
of the errors from the different portions will give a more probable value 
of the error of the experiment. The last two experiments are thus dis- 
cussed in Tables 7 and 8. 49 
Table 7.-Discussion of experiment of March 25, 1897. 
Periods. 
1-4 
(e) 
Heat as 
meas- 
ured. 
Cals. 
375.1 
375.3 
341 6 
Tem 
(f) 
At the 
begin- 
ning. 
°C. 
21.42 
21.44 
21 3R 
perature I 
(g) 
At the 
end. 
°C. 
21.33 
21.29 
21 32 
^o. 5. 
(h) 
Change, 
g -i- 
- 
-0.09 
-0.15 
-0.06 
(i) 
Capacity 
correc- 
tion, 
h X 60. 
Cals. 
-5. 4 
-9.0 
-3.6 
(j) 
Corrected 
heat, 
e + i. 
Cals. 
369.7 
366.3 
338.0 
(k) 
Heat put 
in by the 
current. 
Cals. 
368.1 
366.3 
337.7 
Error in 
heat 
measure- 
ment, 
j-k. 
Cals. 
+ 1.6 
0.0 
+0.3 
2-5 
3-6 
0.6 
= 0.17 per cent. 
357.7 
Sum 
-Avg 
. 1,074.0 
. 358.0 
1, 072.1 
357.4 
+ 1.9 
+0.6 
Table 8.-Discussion of experiment of March 26, 1897. 
Periods. 
(e) 
Heat as 
meas- 
ured. 
Temperature No. 5. 
(i) 
Capacity 
correc- 
tion, 
h X 60. 
(j) 
Corrected 
heat, 
e + i. 
(k) 
Heat put 
in by the 
electric 
current. 
Error 
in heat 
measure- 
ment, 
j -k. 
(f) 
At the 
begin- 
ning. 
(g) 
At the 
end. 
(h) 
Change, 
g -f- 
Cals. 
°C. 
°C. 
°C. 
Cals. 
Cals. 
Cals. 
Cals. 
1-6 
752.0 
22.63 
22. 60 
-0.03 
-1.8 
750.2 
751.5 
-1.3 
2-7 
752.5 
22.61 
22. 60 
-0.01 
-0.6 
751.9 
751.1 
+0.8 
3-8 
751.2 
22.65 
22.64 
-0.01 
-0.6 
750. 6 
750.6 
0.0 
4-9 
750. 2 
22. 67 
22. 62 
-0. 05 
-3.0 
747.2 
749.6 
-2.4 
5-10 
750.6 
22.57 
22.61 
+0. 04 
+2.4 
753.0 
752.2 
+ 0.8 
Sum 
. 3,752.9 
3, 755.0 
-2.1 
0.4 
750.6 
= 0. 053 per cent. 
Avg 
750. 6 
751.0 
-0.4 
In the first column is indicated the number of periods corresponding 
to the section of the experiment taken for discussion. In the second, 
(e), is the heat measured during the same period-i. e., the sum of the 
values in column (e), Tables 4 and 6. The next three columns show the 
change in temperature of the apparatus, and the necessary correction 
is given in the sixth column. Applying these corrections to the values 
in the second column gives the corrected heat in (j). The values of 
the heat generated, (k), are obtained by adding those in column (k), 
Tables 4 and 6, pages 46 and 48. The last column gives the differences 
between the heat generated and that measured, and the average of 
these shows the error of the apparatus. In these experiments the 
error averages about 0.1 per cent, which is as close as the electrical 
energy could be measured. 
Another test of a somewhat different nature was made April 30. A 
small incandescent electric lamp, rated at 4 candlepower, but burned 
low, taking 4.21 watts, or 1 small calorie per second, was placed 
within the chamber and the heat given off was measured. The heat 
17951-No. 63 4 50 
being so small it was necessary to have the water enter at a tempera- 
ture nearly as high as the air inside, and to How in a very slow stream. 
The results of six hours' run are given in Table 9. 
Table 9.-Experiment with 4-candlepower electric lamp-Summary of heat measurements, 
April 30, 1897. 
Time. 
(a) 
Dura- 
tion of 
period. 
Calorimetric measurements. 
Electrical measurements. 
Temper- 
ature of 
the in- 
side by 
the cop- 
per ther- 
mometer 
No. 5. 
(b) 
Water. 
(c) 
Differ- 
ence in 
tempera- 
ture by 
bridge 
No. 6. 
(d) 
Corrected 
difference 
in tem- 
perature, 
c -0.09°. 
(e) 
Amount 
of heat, 
b X d. 
E 
Voltage 
on 
lamp. 
I 
Current 
through 
lamp. 
W 
Watts, 
EXi. 
(k) 
Calories, 
W x a 
X .2378. 
h. m. s 
Secs. 
Kilos. 
°C. 
°C. 
Cals. 
Amp. 
°C. 
9 56 00 
19. 72 
10 11 45 
945 
1 
1.82 
1.73 
1.72 
5.61 
0. 75 
4.21 
0. 95 
19.70 
10 40 35 
1,730 
2 
1.35 
1.26 
2.52 
5. 61 
.75 
4.21 
1.73 
19.65 
10 54 55 
860 
1 
1.28 
1.19 
1.19 
5.61 
.75 
4.21 
.86 
19.63 
11 38 45 
2,630 
3 
1.15 
1.06 
3.18 
5.61 
.75 
4.21 
2.63 
19.62 
1 55 00 
8,175 
9 
.97 
.88 
7.92 
5.61 
. 75 
4.21 
8.17 
19.63 
2 22 30 
1,650 
2 
.41 
.32 
.64 
5.61 
.75 
4.21 
1.65 
19. 66 
3 08 00 
2. 730 
2.6 
.61 
.52 
1.35 
5.61 
.75 
4.21 
2. 73 
19. 72 
3 54 00 
2,760 
3.3 
.95 
.86 
2.84 
5.61 
.75 
4.21 
2.76 
19. 73 
Heat measured 
=21.37 
Heat generated 
=21.48 
It will be seen that the measurement is very close. There is appar- 
ently a small capacity correction, which, if applied, makes the error in 
the opposite direction. It is so small, however, that it is impossible to 
say with certainty how much the correction should be. 
As this amount of heat is only about 1 per cent of the maximum 
which the apparatus will measure and about 3 per cent of the heat 
usually measured, it is evident that there can be no appreciable loss or 
gain of heat by radiation or otherwise. We therefore believe that this 
calorimeter, with proper manipulation, will measure heat as accurately 
as it is possible to measure it by any other form of calorimeter. 
ALCOHOL TEST EXPERIMENTS. 
The electrical tests above described demonstrate the accuracy of the 
apparatus as a calorimeter when the evolution of heat inside the cham- 
ber is reasonably uniform, and there is no current of air passing 
through and no vaporization of water within the chamber. In experi- 
ments with men, however, the development of heat is less uniform; 
furthermore, a current of air is passing through the chamber, water 
and carbon dioxid are produced, and water is vaporized within it. In 
these experiments the principal measurements made with the aid of 
the apparatus are the amounts of water and carbon dioxid produced 
by the man and the heat given off from his body. In experiments 
where the external muscular work has been considerable the latter has 
been transformed into electrical energy, measured, and transformed 51 
into heat. This heat has been measured with that given off directly 
by the body. The crucial test of the accuracy of the determinations 
of carbon dioxid, water, and heat must be made in test experiments 
under circumstances closely similar to those of the actual experiments 
with men. When ethyl alcohol is burned in air carbon dioxid, water, 
and heat are produced. If known quantities of alcohol be burned 
inside the chamber while a current of air is passing through, the con- 
ditions will approach very closely those of an experiment with man. 
To make these experiments accurate, the amount of alcohol burned 
must be exactly known and the combustion must be complete. 
Several series of experiments were made in which alcohol was burned 
inside the chamber. The rate of flow of the ventilating current, the 
method of analysis of the incoming and outgoing air, and the method 
of measurement of heat were the same as if a man were inside the 
chamber. 
COMPOSITION OF ALCOHOL USED-SPECIFIC GRAVITY. 
When ethyl alcohol is diluted with pure water and no other com- 
pounds are present, the quantity of alcohol can be determined with 
reasonable accuracy, even to the fifth or sixth decimal place, by use 
of the pyknometer devised and described by Dr. Edward R. Squibb,1 
who has given more attention to the subject than anyone else of our 
acquaintance. Dr. Squibb has very kindly supplied us with a pyknom- 
eter calibrated by himself. He has also furnished us with a quantity 
of alcohol as nearly pure as it can well be prepared, and also with 
samples diluted with 12 per cent of water. He informs us that, while 
absolute alcohol absorbs moisture from the air with great avidity, a 
solution containing this proportion of water is very slightly altered 
on short exposure to the air. Dr. Squibb also informs us that the high 
grades of commercial ethyl alcohol are so nearly free from other sub- 
stances than ethyl hydroxid and water, and that the composition is so 
slightly changed by the evaporation which takes place with careful 
manipulation, that the determination of the specific gravity by use of his 
pyknometer gives the composition with as great accuracy as is attained 
by combustion with oxygen by the ordinary process of analysis. 
We desire to express our appreciation of the most valuable assistance 
rendered us by Dr. Squibb. 
In tests here made we have used pure ethyl alcohol furnished by Dr. 
Squibb in some cases, but for the larger part of our work, including 
the alcohol tests with the respiration calorimeter, a high grade of 
commercial alcohol was used instead of the so called absolute alcohol, 
because it can be more easily stored, weighed, and measured without 
change of water content. The specific gravity of this alcohol was deter- 
mined at the time of each experiment. The variations in the different 
samples were small, the proportion of alcohol being not far from 90.5 per 
'Jour. Amer. Chern. Soc., 19 (1897), p. Ill; Ephemeris, 2 (1884-85) p. 541. 52 
cent by weight. From the percentage of alcohol present the amount of 
carbon dioxid and water that would be formed by complete oxidation 
is readily calculated as follows: One gram of ethyl alcohol (C2H6O) 
will yield on oxidation 1.9110 grams of carbon dioxid and 1.1737 grams 
of water.1 One gram of a mixture of alcohol and water containing 
90.77 per cent absolute alcohol (the strength of that used in the first 
three tests) will give 1.9110x0.9077, or 1.7346 grams of carbon dioxid, 
and 1.1737 x 0.9077, or 1.0654 grams of water from the combustion of the 
ethyl alcohol in addition to the 0.0923 grams of water present in the 
mixture, making a total of 1.1577 grams of water from 1 gram of the 
alcohol mixture used. The estimates of the quantity of heat produced 
by the oxidation of a gram of each of the different specimens of alcohol 
are made by use of the factor 7.067 calories for the heat of the combus- 
tion of 1 gram pure ethyl alcohol as explained beyond. The strength 
of the alcohol used in each test, and the factors for the computation of 
the amounts of carbon dioxid, water, and heat given off per gram, are 
shown in Table 10. 
Table 10.-Strength of alcohol used in the different tests and the corresponding amounts 
of carbon dioxid, water, and heat yielded by the combustion of 1 gram. 
Test No. 
Ethyl 
hydroxid 
Products of oxidation of 
1 gram. 
by 
weight. 
Carbon 
dioxid. 
Water. 
Heat. 
Per cent. 
Gramt. 
Grams. 
Sm. cals. 
1 
90. 77 
1.7346 
1.1577 
6,415 
2 
90.77 
1. 7346 
1.1577 
6, 415 
3 
90.77 
1.7346 
1. 1577 
6,415 
4 
90.61 
1.7816 
1.1574 
6, 403 
5 
90.63 
1.7319 
1.1574 
6. 405 
6 
90.26 
1.7249 
1.1567 
6. 379 
90.26 
1.7249 
1.1567 
6,379 
8 
90.61 
1.7316 
1.1574 
6,40 i 
9 
90. 26 
1.7249 
1. 1567 
6. 379 
HEAT OF COMBUSTION OF ALCOHOL-DETERMINATION WITH THE 
BOMB CALORIMETER. 
The heat of combustion of the alcohol used in the tests was deter- 
mined by the bomb calorimeter.2 Two methods were used. In one the 
'Using the values 0 = 16, C = 12, H = 1.008 for atomic weights, as proposed by the 
committee on atomic weights, of the American Chemical Society, Jour. Amer. Chern. 
Soc., 20 (1898) p. 163. See also Richards, Amer. Chern. Jour., 20 (1898), p. 543; also the 
values proposed by the committee of the German Chemical Society, Ber. Deut. 
Chern. Gesell.,31 (1896), p. 2761. 
2The form of apparatus used, which is a modification of that of Berthelot, is 
described in U. S. Dept. Agr., Office of Experiment Stations Bui. 21, and Connecticut 
Storrs Sta. Rpt. 1897. The determinations of the heats of combustion of alcohol 
used for these tests and in other experiments connected with the general investiga- 
tion were made by Dr. O. F. Tower and the details of the work are not yet published. 53 
alcohol was poured upon cellulose absorption blocks.1 The weighing 
was made with due precautions to prevent, if possible, any loss by 
evaporation. The alcohol and absorption blocks were burned in the 
calorimeter, a correction being applied for the heat of combustion of the 
absorption block used. Eight determinations were made by this method 
with commercial alcohol of high grade and of different strengths. 
The use of absorption blocks is objectionable because of the difficulty 
of avoiding evaporation of alcohol during the process of weighing and 
transferring to the bomb, and because of the considerable and not 
absolutely certain correction to be applied for the heat of combustion 
of the absorption blocks used. The first difficulty was avoided entirely 
and the other partially by the second method employed. In this the 
alcohol was inclosed in small gelatin capsules by which the evaporation 
is prevented. The weight of gelatin is also small and the correction for 
its heat of combustion is less than with the absorption blocks. Eight 
determinations were made with Squibb's alcohol and nine with com- 
mercial alcohol, such as was used in the tests in the respiration calo- 
rimeter. The results obtained by both methods are shown in Table 11. 
Table 11.-Determination of heats of combustion of alcohol in the bomb calorimeter. 
Ethyl 
hydroxid 
by 
weight. 
Heat of 
combus- 
tion per 
gram ac- 
tually de- 
termined. 
Calcula- 
ted heat 
of com- 
bustion 
per gram 
of ethyl 
hydroxid. 
Alcohol contained in absorption blocks: 
Per cent. 
Small cals. 
Small cals. 
Specimen No. 1 
89.9 
6,348 
7, 062 
Do 
89.9 
6,345 
7, 058 
Do 
89.9 
6,377 
7, 093 
Average of above 3 
7, 071 
Specimen No. 2 . - 
81.3 
- 
5, 744 
7, 063 
Do 
81.3 
5,763 
7, 090 
7,076 
Average of above 2 - - 
72.5 
- 
Specimen No. 3 
5,132 
7,078 
Do 
72.5 
5, 094 
7,030 
Do 
72.5 
5,132 
7,078 
Average of above 3 
7, 062 
Average of above 8 
7, 069 
Alcohol contained in gelatin capsules: 
Specimen No. 4 
88.0 
6, 205 
7, 050 
Do 
88.0 
6,224 
7, 072 
Do 
88.0 
6,209 
7,055 
Do 
88.0 
6,199 
7, 044 
Do 
88.0 
6,204 
7,050 
Do 
88.0 
6, 203 
7,049 
Do 
88.0 
6, 227 
7,076 
Do 
88.0 
6, 240 
7, 090 
Average of above 8 
7, 061 
""-" -- 
- 
- 
'As described by Kellner, Landw. Vers. Stat., 47 (1896), p. 297. 54 
Table 11.- Determination of heats of combustion of alcohol, etc.-Continued. 
Ethyl 
hydroxid 
by 
weight. 
Heat of 
combus- 
tion per 
gram ac- 
tually de- 
termined. 
Calcula- 
ted heat 
of com- 
bustion 
per gram 
of ethyl 
hydroxid. 
Alcohol contained in gelatin capsules-Continued. 
I'er cent. 
Small cals. 
Small calt. 
Specimen No. 5 
90.61 
6,435 
7, 101 
Do 
90. 61 
6, 384 
7, 046 
Do 
90.61 
6,401 
7, 064 
Do 
90. 61 
6, 383 
7,045 
Do 
90.61 
6,403 
7,066 
Do 
90.61 
6, 397 
7, 060 
Do 
90.61 
6, 436 
7, 103 
Do 
90. 61 
6, 433 
7, 099 
Do 
90. 61 
6, 386 
7. 047 
Average of above 9 
7, 070 
Average of above 17 
7, 066 
Average of above 25 
7,967 
The heat of combustion when determined by the first method ranged 
from 7,030 to 7,090, and averaged 7,069 small calories per gram of abso- 
lute alcohol. 
The results obtained by the second method with Squibb's alcohol 
ranged from 7,044 to 7,090 small calories, and averaging 7,061 small 
calories per gram, absolute alcohol. The results of similar determina 
tions with the commercial alcohol ranged from 7,045 to 7,103 small 
calories, and averaging 7,070 small calories per gram, absolute alcohol. 
The average of these 17 determinations with alcohol in gelatin capsules 
gives 7,066 small calories, practically the same result as was obtained 
by the use of the absorption blocks. Considering the range of variation 
in these different determinations, we would hardly be justified in assum- 
ing that the figure 7,067 represents exactly the heat of combustion of 
ethyl hydroxid. There is reason to hope, however, that some of the 
sources of error in the determinations may be partially eliminated, 
and that thus more reliable results may be obtained. It is worthy of 
note that Berthelot and Matignon1 obtained the figure 7,068 as the 
average of two determinations, which gave 7,067.3 and 7,068.5 small 
calories, respectively, at 13°. The average 7,068 corresponds to 7,079 
at 20°, a value very close to those obtained here. 
The heat of combustion of alcohol of any given dilution with water 
is found by multiplying the heat of combustion of 1 gram of absolute 
alcohol by the percentage present in the specimen. Thus in the speci- 
men of alcohol used in the first three tests the heat of combustion was 
7,067 x .9077, or 6,415 small calories per gram. 
While the heat of combustion of 1 gram of absolute alcohol is thus 
7,067 small calories per gram, this does not represent the amount of 
'Ann. Chim. et Phys., 6. ser., 27 (1892), p. 312. 55 
heat that is given off by the combustion of 1 gram of alcohol within 
the chamber of the respiration calorimeter. In the bomb calorimeter 
all the water vapor formed is condensed within the apparatus, and 
hence the heat that had been required to vaporize the water is given 
off again. In the respiration calorimeter, on the other hand, the water 
passes out as vapor in the ventilating air current. The heat required 
to vaporize it comes from the combustion of the alcohol, and is not 
measured by the calorimeter. In order to obtain the total amount of 
heat given off in the combustion, therefore, the heat actually measured 
must be added to the amount required to vaporize the excess of water 
in the outgoing over that in the incoming air current, which is the 
amount of water vaporized at the expense of the heat produced in the 
combustion of the alcohol. 
THE CALORIE HERE USED AS THE UNIT OF MEASURE. 
Although the specific heat of water is often taken as unity for all 
temperatures, it actually varies by an appreciable amount. Hence if 
the unit of heat be defined as the amount required to raise unit mass 
of water 1 degree, this unit will be a variable one. The theoretical 
large calorie, namely, the quantity of heat that will raise a kilogram of 
water from 0° to 1°, or from 4° to 5°, is a very inconvenient unit in 
practice. Moreover, the specific heat of water changes quite rapidly at 
these low temperatures, and hence a higher temperature is more favor- 
able for the working unit. Many authorities take the specific heat of 
water at 15° C. as unity. Inasmuch, however, as we have used 20° C. 
as our standard temperature for the respiration calorimeter in experi- 
ments with human subjects, and inasmuch also as the specific heat of 
water has a minimum near to 20° and changes very slowly in that 
vicinity, we have found it desirable to take as our working unit of heat 
the calorie at 20°. In other words, the large calorie at 20°, which we 
designate in the tables as C20, is the amount of heat required to raise 
the temperature of a kilogram of water 1 degree at 20° C. (that is, from 
half a degree below 20° to half a degree above). The large calorie at 
any other temperature is here designated by Ct, and is the amount of 
heat required to raise the temperature of a kilogram of water 1 degree 
at the temperature t. 
Since the specific heat of water is nearly a minimum at 20° C., our 
standard calorie, C2(,? is nearly always less than Ct. This difference is 
ordinarily neglected. In experiment No. 6 with a man described 
beyond, in which work was performed, it might perhaps be neglected, 
since it is considerably less than some of the other errors of the experi- 
ment. But in the alcohol-test experiments and in the experiment with 
a human subject in which no work was performed, this difference is, 
relatively speaking, not inconsiderable. In Table 12 the specific heat 
of water is given for different temperatures between 0° and 31°. The 56 
figures are based upon the experiments of Rowland,1 Bartoli and Strac 
ciati,1 Griffiths,1 and Ludin.2 The differences in the results obtained 
by these investigators at temperatures below 22° are very small, and, 
in view of the care with which their experiments were made, we do not 
believe that the estimates of this table can be far enough from the 
truth materially to diminish their value for the present purpose. 
Table 12.-Specific heat of water at different temperatures referred to that 
at 20° C. as unity. 
Tempera- 
ture ° C. 
Specific 
heat. 
Tempera- 
ture ° C. 
Specific 
neat. 
0 
1.0090 
16 
1.0007 
1 
1.0083 
17 
1.0004 
2 
1.0076 
18 
1.0002 
3 
1.0069 
19 
1.0001 
4 
1.0062 
20 
1.0000 
5 
1.0056 
21 
. 9999 
6 
1.0050 
22 
.9998 
7 
1.0044 
23 
.9998 
8 
1. 0039 
24 
.9998 
9 
1.0034 
25 
.9998 
10 
1.0029 
26 
.9998 
11 
1.0024 
27 
.9999 
12 
1.0020 
28 
.9999 
13 
1.0016 
29 
1.0000 
14 
1.0013 
30 
1.0001 
15 
1.0010 
31 
1.0002 
It will be seen that at 0° the calorie (Co) is nearly 1 per cent greater 
than C20, and that C]o is about three parts in a thousand greater than 
G^. 
The results of all the combustions by the bomb calorimeter, as well 
as the measurements by the respiration calorimeter, are to be expressed 
in terms of O^. To do this it is necessary to know the mean specific 
heat of water for the range of temperature employed in any given 
experiment. For example, if water is warmed from 3° to 17° in pass- 
ing through the absorbing pipes of the respiration calorimeter, the 
result will be in terms of Cu-m; that is, in terms of the mean calorie 
from 3° to 17°. This is 1.0032 times as great as the standard calorie C20, 
whereas Ci0, the calorie for the mean temperature, is only 1.0029 times 
as great as C20. In other words, the mean specific heat from 3 to 17° 
is 1.0032 times the specific heat at 20°, whereas the specific heat at 
10°, the mean temperature, is 1.0029 times that at 20°. The difference 
between the mean specific heat for this range of temperature and the 
specific heat of the mean temperature is therefore appreciable; this is 
1 See p. 43. 
3Inaug. Diss., Zurich, 1895; cited by Longuinine, Bestimmung der Verbreunungs- 
wiirme, Berlin, 1897, p. 17. See discussion of results of experiments on the specific 
heat of water on pp. 12-20 of this valuable treatise. 57 
of course because the variation of the specific heat is not linear. We 
have accordingly ejaculated the mean specific heat of the water in every 
case for the range of temperatures employed, and expressed the heat 
measured in terms of C20. In the tables this range of temperature is 
given in the fourth column, and the heat in terms of C20 in the fifth 
column. 
THE LATENT HEAT OF WATER VAPOR. 
The value commonly used for the latent heat of vaporization of water 
at different temperatures is that given by Regnault1 as the result of 
his classical investigations half a century ago. Regnault's method, for 
temperatures between 63° and 195°, consisted in condensing, within a 
suitable receptacle placed inside a water calorimeter, the vapor of 
water generated by ebullition in a separate boiler, the pressure in 
which could be varied from a small fraction of 1 atmosphere to 20 at- 
mospheres or more. The quantity of heat yielded by the vapor upon 
condensation was determined by the increase of temperature of the 
water surrounding the condenser. The apparatus was large and very 
elaborate, and the investigation was carried out in a masterly manner. 
Regnault's formula, which expresses quite closely the results obtained 
throughout the given range of temperatures, i. e., from 63° to 195°, is 
H= 606.5+.305 t. H represents what Regnault called the "total heat" 
of steam or water vapor, at the temperature f; that is, it is the number 
of small calories required to raise a gram of water from 0° C. to any tem- 
perature t and then completely vaporize it at that temperature. For 
<=100°, this gives H=637. The "latent heat of vaporization" at any 
temperature, as distinguished from the "total heat" at the same tem- 
perature, is the number of units of heat required to evaporate the water 
after it has been brought to the temperature of evaporation. Hence at 
100° C. the latent heat of vaporization is 537 calories. The formula for 
the latent heat at any temperature t is 
L=H-< 
L=606.5+.305 t-t 
or L=606.5 -.695 t 
Regnault's value for the latent heat of steam at 100° is abundantly 
confirmed by later researches. Andrews, Favre and Silberman, Berthe- 
lot and others obtain results very closly agreeing, and the accuracy of 
the determinations leaves little to be desired so far as this temperature 
is concerned. 
At lower temperatures than 63°, however, this method was inappli- 
cable because the change of state from water to steam at low pressure 
is irregular and explosive; hence a different method became neces- 
sary. The process was accordingly reversed, and water contained in a 
small reservoir inside a much smaller calorimeter than that previously 
1 Mem. Acad. Roy. Sci. Inst. France, 21 (1847), pp. 635-728. 58 
used was evaporated at reduced pressure. The heat absorbed in the 
vaporization of the water (about 5 grams at each experiment) was 
then determined from the lowering of the temperature of the water which 
surrounded the vessel in which the evaporation took place. The 
results, as Regnault himself points out, were subject to comparatively 
large experimental errors, and were very discordant. The mean results 
over a range of temperature from 0° to 16° are fairly well represented 
by the formula given above for higher temperatures. 
In the hope of obtaining better results at these low temperatures 
Regnault tried a third method.1 This consisted in passing a current 
of air through a specially arranged receptacle containing water, and 
so carrying away the water vapor in air at external atmospheric 
pressure. The heat absorbed was determined, as before, by the low- 
ering in temperature of the surrounding water. Correction was of 
course made for the heat yielded or absorbed by the air current itself, 
according as it left the calorimeter cooler or warmer than upon enter- 
ing it. The aii1 was not entirely saturated, and hence the circum- 
stances of this experiment corresponded very closely to those under 
which water vapor is carried away from the interior of our respiration 
calorimeter. Regnault made only two series of experiments by this 
method, there being four experiments in the first series and three 
in the second. Unfortunately for our present purpose, the results 
obtained were very unsatisfactory. The mean of the first series gave 
11 = 612.8 for 16.5°, instead of 611.5 which the formula requires; 
and the mean of the second series gave 623.2 at 18°, instead of 
612 as required by the formula. The individual determinations of the 
first series were 668.5, 616.9, 608.7, 617.2, thus presenting very large 
differences. 
One reason for so large discrepancies appears to be the large cooling 
corrections to which the apparatus was subject. For example, in the 
first experiment the temperature of the water of the calorimeter 
dropped from 17.33° to 13.95°, a fall of 3.38°. But after applying the 
necessary corrections due to disturbing influences, the difference is 
reduced to 2.5873°. That is, about 25 per cent of the lowering of 
temperature was due to external disturbing causes, and it is hardly 
to be expected that such large corrections can be determined with 
sufficient accuracy. 
Moreover, Regnault appears to have made a serious error in com- 
puting his results. In the formula given2 X and t are inadvertently 
interchanged. Making this correction, the formula expresses the value 
of the total heat in terms of the various quantities determined by the 
experiment. In the tabulated results,3 the total heat is recorded as the 
latent heat of vaporization, there called A and then the "heat of the 
1 Acad. Roy. Sei. Inst. France, 26 (1862), pp. 883-890. 
2 Ibid., p. 886. 
3 Ibid., pp. 888, 889. 59 
water," numerically equal per gram to the mean temperature of the 
wsiter at evaporation, is added again to give the values recorded as the 
total heat, called X. If the values of A for the total heat are taken, as 
it appears to us should be done, the value for the latent heat of vapori 
zation of water obtained by this third method are smaller than by the 
second method and smaller than the formula gives. However, the 
results are too discordant to be of much value, whether X or A be taken 
as the total heat. 
Starkweather,1 in a critical review of various determinations of the 
latent heat of water vapor, quotes the work of Dieterici,2 Griffiths,3 
and Svensson4 as the best that has been done at low pressure and 
temperatures. Dieterici and Svensson determined with an ice calori- 
meter the latent heat of water vapor at 0° C. Dieterici's work is very 
carefully done, and >his result is L = 598.9 at 0° C., in terms of the 
quantity of heat required to raise 1 gram of water from 15° to 16° as 
the unit of heat. Svensson's result agrees very closely with this, 
being 599.9. Griffiths measured the latent heats of water vapor be- 
tween 25° and 50° C., and succeeds in representing them by the formula 
L = 596.73 - .6011. Extrapolating to 0° and 100°, this agrees quite well 
with Dieterici at 0° and Regnault and others at 100°, being 596.73 at 
0° and 536.6 at 100°. • None of these experiments, however, have been 
carried out at the temperature commonly used in the respiration calori- 
meter-i. e., 20° C. Hence we are obliged to deduce the latent heat of 
vaporization at 20° from formulas given by experiments at other tem- 
peratures. 
Esing Regnault's original formula, from which L = 606.5 - .695 t we 
get L = 592.6 at 20° C. Using Starkweather's modification of this for- 
mula, which agrees with Regnault's results at temperatures between 
63° and 100° better than Regnault's, viz, L = 598.9 - .5581 - .00064 t2, 
L becomes 587.5 at 20° C. Using Griffiths's formula above as quoted 
L = 584.5 at 20° C. 
Some preliminary experiments have been made with the respiration 
calorimeter, with a view of determining this quantity L, i. e., the latent 
heat of vaporization of water at 20°, under the circumstances of the 
respiration experiment. The results so far have not, however, given a 
satisfactory value for this quantity. We have therefore taken 592 as 
a provisional value for the calculation of the experiments reported in 
this bulletin, and it would appear that this is probably within 1 per 
cent of the truth, although it is possibly as much as 2 per cent in error. 
Further investigations at 20° are necessary in order to fix the value 
more precisely. 
'"Concerning Regnault's calorie and our knowledge of the specific volumes of 
steam." Amer. Jour. Sci., 7 (1899), p. 13. 
2Ann. Phys. u. Chern. [Wiedemann], 38, 1889. 
3Phil. Trans. Roy. Soc. [London], A, 1895. 
4 Beiblatter, Ann. Phys. u. Chern. [Wiedemann], 20, p. 356. 60 
In a preliminary publication 1 of results of these experiments the total 
heat of water vapor at 20° (i. e., 612) was taken by an oversight instead 
of the latent heat in computing the amount of heat carried away by the 
water vapor in the ventilating current of air. But inasmuch as the 
temperatures in the calculations for these experiments are all reduced 
to 20° C. instead of 0°, the heat above 20° should be used, which is 
of course the latent heat itself when the water vapor is carried away 
at 20o c. 
ARRANGEMENTS FOR THE COMPLETE COMBUSTION OF ALCOHOL IN 
THE TEST EXPERIMENTS. 
Various substances have been used by experimenters with different 
forms of respiration apparatus for burning within the chamber in order 
to deliver known quantities of carbon dioxid and water for the purpose 
of testing the accuracy of the apparatus and methods. The chief diffi- 
culty is to find substances which can be so burned as to insure complete 
oxidation of the carbon and hydrogen. The first attempts in this lab- 
oratory with the combustion of alcohol were unsatisfactory because the 
oxidation of the alcohol was incomplete, and a considerable amount of 
experimenting was necessary in order to learn the conditions under 
which complete combustion could be secured. As the result a small 
lamp, such as is ordinarily used with kerosene for illuminating pur- 
poses, has been employed. By proper arrangement of wick and chim- 
ney the alcohol is burned so completely that no traces of volatilized 
alcohol, acetone, aldehyde, carbon monoxid, or other substances capable 
of yielding carbon dioxid upon heating with oxygen could be found 
among the products of combustion.2 
DESCRIPTION OF THE TEST EXPERIMENTS. 
A considerable number of alcohol test experiments were made. The 
individual experiments included from one to seven periods or "runs" of 
approximately six hours each. Three of these tests are here described in 
detail in order to show how they were made and how the results were 
calculated. These three tests were made in the spring of 1897, each 
being followed by an experiment with a man in the chamber. The 
" work" experiment with a man, No. 6, described beyond, pages 74 to 87, 
was made between the second and third of these tests. 
The lamps used in these tests contain approximately 380 grams of 
1 Connecticut Storrs Sta. Rpt. 1897, p. 212. 
2 Jour. Amer. Chem. Soc., 20 (1898) p. 293. The details of a considerable amount of 
experimental data accumulated upon this subject by Drs. F. G. Benedict and W. J. 
Karslake are reserved for future publication. It will suffice to say that a large 
number of preliminary experiments were made with various organic compounds of 
known purity and composition, and alcohol was selected because of the ease of 
securing complete combustion under proper conditions. The lamp and arrange- 
ments for burning the alcohol inside the chamber were the same as in the experi- 
ments where the combustion was proved to be complete. 61 
alcohol. By adjusting the wick the rate of burning was so regulated 
as to give off carbon dioxid and water in the desired amounts. Under 
no circumstances was the flame allowed to have a luminous tip. Just 
before the beginning of each six-hour period of the test the lamp was 
filled, lighted, and the wick adjusted. The flame was then extinguished. 
After the lamp had cooled it was weighed, immediately put into the 
chamber, and lighted. The lamp was introduced into the chamber 
through the food aperture and was placed in a basket suspended from 
the ceiling of the chamber. When a change was made from one lam]) 
to the other between the periods the ventilating air current was stopped, 
the cover removed from the food aperture, and the basket containing 
the lamp drawn close to the edge of the aperture by one attendant, 
while another reached in and removed the lamp, extinguishing it as 
soon as it was brought out. Another lamp was then placed in the 
basket, lighted, and the basket allowed to swing back to its normal 
position, the cover replaced on the food aperture, and the air current 
started again. The whole operation of removing one lamp and intro- 
ducing the other, including shutting off and starting the air current, 
usually required less than a minute. The lamp that was removed was 
weighed at once, and the loss of weight was taken as the measure of 
the amount of alcohol burned during the period. In computing the 
amounts of heat given off" by the combustion of the alcohol, allowance 
was made for the heat brought out of the chamber in the heated lamp 
chimney. 
The lamp itself and its contents were not heated appreciably, except- 
ing the metal burner, and the amount of heat carried away by the lat- 
ter was very much less than by the chimney. Several determinations 
of the heat of the chimney were made by dropping it into the water 
contained in a calorimeter and noting the rise of temperature pro- 
duced. The results ranged from about 1.5 to 2.5 large calories. 
Accordingly, the following estimates, which make a small allowance 
for the heat of the burner, are used as corrections in computing the 
results: For the maximum rate of burning, 2.5 large calories; for the 
mean rate of burning, 2 large calories, and for the minimum rate of 
burning, 1.5 large calories. 
One of these numbers is to be added to the quantity of heat measured 
in each experimental period. This correction amounts to about three 
parts in a thousand of the heat measured. While this correction is 
itself not entirely satisfactory, the error can hardly be large enough to 
materially affect the result. 
In Table 13 are shown the date and duration of each experiment and 
each period; the total amount of alcohol burned and the rate of com- 
bustion per hour; the number of liters of air in the ventilating air cur- 
rent; the data obtained for determining the amounts of carbon dioxid 
and water produced by the burning of the alcohol, and the comparison 
of the amounts thus determined with the theoretical amounts produced 62 
by the oxidation of the same amounts of alcohol. The number of milli- 
grams per liter of carbon dioxid in the incoming and in the outgoing 
air current, as sampled by the aspirators, is shown in the sixth and 
seventh columns. The difference between these quantities represents 
the excess in the outgoing air current-i. e., the amount added by the 
combustion of the alcohol to each liter of air in the ventilating air cur- 
rent. This excess per liter multiplied by the number of liters gives 
the total excess. The total amounts thus added are shown in the eighth 
column of the table. Similar data serve for the computation of the 
total excess of water in the outgoing air current not removed by the 
freezers. The amount of water condensed in the freezers added to the 
excess gives the total excess in the outgoing air current. This total 
excess is not shown in the table. 
If the amounts of carbon dioxid and water in the air in the chamber 
at the end of a given experimental period are not the same as at the 
beginning, allowance must be made for the gain or loss of these com- 
pounds in calculating the amount produced by the combustion of the 
alcohol during the period. When the rate of combustion of the alcohol 
is quite constant, there is but little need of analyses of the residual air 
in the chamber at the beginning and end of each experimental period. 
In such an experiment as No. 3, however, where the rate of combus- 
tion was purposely made irregular in order more nearly to approach to 
the condition existing in experiments with men, analyses of the resid- 
ual air are necessary. The ninth and sixteenth columns of Table 13 
show the amounts of carbon dioxid and water found in the residual 
air of the chamber. Columns 10 and 17 of the table show the amounts 
actually measured, these amounts being corrected for differences in the 
residual amounts in the chamber when such determinations were made. 
With these values the theoretical amounts given off* by the combustion 
of the alcohol during the different periods are compared. The results 
of the three tests reported in full are shown in the following table. 63 
Table 13.-Amounts of carbon dioxid and water actually obtained by the combustion of different quantities of alcohol in the respiration calorimeter as 
compared with the theoretical amounts from the same quantities of alcohol. 
Tests and periods. 
Alcohol burned. 
Carboi 
dioxid. 
Water. 
V entila- 
ting air 
current. 
In outgoing air. 
Resid- 
Deter- 
Theoret- 
In outgoing air. 
Resid- 
Deter- 
Theoret- 
In in- 
ual 
mined 
ical 
In in- 
Excess 
not con- 
densed 
in freez- 
ers. 
ual 
mined 
ical 
Number ana date. 
Dura- 
tion. 
Total. 
Per 
hour. 
coming 
air per 
Per 
Total 
amount 
in 
amount 
from 
amount 
from 
coming 
air per 
Per 
Con- 
densed 
amount 
in 
amount 
from 
amount 
from 
liter. 
liter. 
excess. 
chain- 
alcohol 
alcohol 
liter. 
liter. 
in freez- 
cham- 
alcohol 
alcohol 
ber. 
burned. 
burned. 
era. 
ber. 
burned. 
burned. 
- 
: 
- 
- 
- 
- 
- 
- 
- 
Test 1. 
1897. 
h. 
m. 
Grams. 
Grams. 
Liters. 
Mgs. 
Mgs. 
Grams. 
Grams. 
Grams. 
Grams. 
Mgs. 
Mgs. 
Grams. 
Grams. 
Grams. 
Grams. 
Grams. 
Period 
1, Apr. 26.... 
6 
0 
120.3 
20.05 
25, 024 
0. 673 
9.035 
209. 24 
209.24 
208. 67 
0.853 
1.038 
4.63 
139.6 
144. 23 
139.27 
Period 
2, Apr. 27.... 
5 
21 
101.3 
18. 95 
21,915 
.839 
8. 707 
172.43 
172.43 
175.72 
.951 
.936 
-0.33 
116.5 
116.17 
117.28 
Period 
3, Apr. 27.... 
6 
3 
112.2 
18.54 
25, 024 
.675 
8.604 
198.43 
198.43 
194.63 
.893 
. .967 
1.85 
131.4 
133. 25 
129.89 
Period 
4, Apr. 27....' 
5 
58 
104.3 
17.50 
25, 024 
.562 
8.019 
186. 60 
186. 60 
180.92 
.838 
.914 
1.90 
121.5 
123. 40 
120.75 
Period 
5, Apr. 27.... 
6 
7 
103. 9 
17.00 
25,024 
.642 
7.758 
178. 07 
178.07 
180.23 
.885 
.854 
- 0. 77 
120.5 
119.73 
. 120.28 
Period 
6, Apr. 28.... 
7 
0 
119.3 
17. 04 
28,911 
.566 
7.657 
205. 00 
205. 00 
206.94 
.828 
.902 
2.10 
135.5 
137. 60 
138.12 
Period 
7, Apr. 28.... 
4 
17 
71.0 
16. 58 
18,176 
.549 
7. 515 
126. 60 
126.60 
123.15 
.883 
1.032 
2. 70 
81.4 
84.10 
82.20 
Period 
8, Apr. 28.... 
5 
56 
119.2 
20.10 
25,175 
.565 
8.482 
199.20 
199.20 
206.76 
.862 
.972 
2. 77 
127.6 
130.37 
138.00 
Period 
9, Apr. 28.... 
6 
2 
117.8 
19. 53 
25,953 
.614 
8.513 
204. 93 
204.93 
204.34 
.789 
.966 
4.53 
134.0 
138.53 
136.38 
Period 10, Apr. 29.... 
5 
55 
111.8 
18.90 
25,174 
.644 
8.305 
192.86 
192. 86 
193.93 
.701 
.843 
3.55 
128.8 
132.35 
129.43 
Period 11, Apr. 29.... 
5 
13 
95.9 
18.40 
21, 260 
.721 
8.517 
165. 95 
165. 95 
166. 35 
.897 
1.003 
2.24 
108.1 
110. 34 
111.02 
Total, 1 to 11... 
63 
52 
1,177.0 
266, 660 
2, 039.31 
2,039.31 
2,041.64 
25.17 
1, 344. 9 
1, 370. 07 
1, 362. 62 
Total, 3 to 11.... 
52 
31 
955.4 
219, 721 
1, 657. 64 
........ 
1, 657. 64 
1, 657.25 
20.87 
1, 088.8 
........ 
1,109. 67 
1,106.10 
, 
■ 
■ 
■ ■ - 
Test 3. 
1897. 
Period 1, May 10 
5 
56 
157.9 
26. 62 
25, 236 
.751 
11. 650 
275. 27 
275. 27 
273.90 
1.208 
1.329 
3.03 
182.6 
185. 63 
182.80 
Period 2, May 10 
6 
0 
147.7 
24. 62 
24, 769 
.757 
11.190 
258.42 
258.42 
256. 20 
1.172 
1.320 
. 3.66 
174/6 
178. 26 
170. 99 
Period 3, May 10 
6 
0 
160.8 
26. 80 
24, 802 
1.132 
12.168 
273.71 
273.71 
278. 90 
1.401 
1.511 
2. 70 
179.9 
182. 60 
186.16 
Period 4, May 11 
6 
0 
165.8 
27.63 
25, 284 
.626 
12. 052 
288. 88 
288. 88 
287. 60 
1.341 
1.532 
4.80 
184.7 
189. 50 
191. 95 
Period 5, May 11 
6 
0 
166.6 
27. 77 
24. 919 
.680 
12. 241 
288.16 
288.16 
288. 98 
1.311 
1.534 
5. 47 
183.5 
188. 97 
192.87 
Total 
29 
56 
798.8 
125,010 
1, 384. 44 
1, 384. 44 
1, 385. 58 
19. 66 
905.3 
924. 96 
924.77 
- 
: 
■ 
~ - ■ - 
===== 
- 
- 
- 
- 
' ■ = 
11 ■ 1 " = 
■ ~= 
===== 
== 
, - ■ == 
= -= 64 
Testa and periods. 
Alcohol 
burned. 
Ventila- 
ting air 
current. 
Carbon dioxid. 
Water. 
Number and date. 
Dura- 
tion.' 
Total. 
Per 
hour. 
In in- 
coming 
air per 
liter. 
In outgoing air. 
Resid- 
ual 
amount 
in 
cham- 
ber. 
Deter- 
mined 
amount 
from 
alcohol 
burned. 
Theoret- 
ical 
amount 
from 
alcohol 
burned. 
In in 
coming 
air per 
liter. 
In outgoing air. 
Resid- 
ual 
amount 
in 
cham- 
ber. 
Deter- 
mined 
amount 
from 
alcohol 
burned. 
Theoret- 
ical 
amount 
from 
alcohol 
burned. 
Per 
liter. 
Total 
excess. 
Per 
liter. 
Excess 
not con- 
densed 
in freez- 
ers. 
Con- 
densed 
in freez- 
ers. 
Test 3. 
1897. 
h. 
m. 
Grams. 
Grams. 
Liters. 
Mgs. 
Mgs. 
Grams. 
Grams. 
Grams. 
Grams. 
Mgs. 
Mgs. 
Grams. 
Grams. 
Grams. 
Grams. 
Grams. 
J20.41 
1 
j43. 91 
Period 1, May 26 
7 
4 
72.5 
10.26 
26,953 
.459 
4.822 
117. 62 
29. 37 
J 126.58 
125. 76 
.907 
1.212 
8.22 
98.8 
131.68 
} 94.79 
83.93 
Period 2, May 26 
5 
0 
94.6 
18.92 
20, 339 
.468 
8. 390 
161.14 
40.32 
172.09 
164.10 
.918 
1.282 
7.42 
102.5 
26. 40 
104.64 
109. 53 
Period 3, May 26 
5 
6 
108.0 
21.18 
18, 846 
.613 
9.949 
175. 95 
49.53 
185.16 
187.34 
1.013 
1.387 
7.05 
109.5 
37. 77 
127. 92 
125.03 
Period 4, May 26 
6 
30 
91.3 
14. 04 
25, 616 
.614 
7.618 
179.46 
34.51 
164. 44 
158. 37 
.897 
1.270 
9. 57 
119.2 
28.85 
119. 85 
105. 70 
Period 5, May 27 
6 
1 
68.4 
11.38 
23,602 
.544 
5.927 
127. 05 
27.41 
119.95 
118.65 
.830 
1.191 
8.52 
86.9 
24.09 
90.66 
79.18 
Period 6, May 27 
6 
9 
70.6 
11.48 
23, 221 
.533 
5.776 
121.77 
25.20 
119. 56 
122.47 
.753 
1.121 
8.54 
83.5 
22. 08 
90.03 
81.73 
Total 
35 
50 
505. 4 
138, 577 
882. 99 
887. 78 
876. 69 
49.32 
600.4 
627.89 
585. 10 
Table 13.-Amounts of carbon dioxid and water actually obtained by the combustion of different quantities of alcohol, etc.-Continued. 65 
The method of applying the data obtained by the analyses of residual 
air is briefly explained on page 38. For example, the quantities given 
in the ninth and sixteenth columns represent the amounts of carbon 
dioxid and water remaining in the apparatus at the beginning of the 
experiment and at the end of each period. There were 20.41 grams 
of carbon dioxid in the air of the chamber at the beginning of the 
first period of test 3, and 29.37 grams at the end. In other words 
(29.37-20.41 = ) 8.96 grams of the carbon dioxid given off by the com- 
bustion of the alcohol during the period remained in the chamber at 
the close and must be added to the total excess of 117.62 grams found 
in the outgoing air in order to obtain a true measure of the amount 
produced in the chamber during this period, which is 126.58 grams. 
In a similar manner wre find that during the fourth period the quan- 
tity of carbon dioxid in the chamber diminished by 15.02 grams, and 
this amount must therefore be deducted from the 179.46 grams total 
excess in the ventilating air current to obtain the determined amount 
given off by the combustion of the alcohol. 
Similar calculations serve to show the amount of water vapor that 
was added to or removed from the air of the chamber during each 
period, and the corresponding amount that must be added to or deducted 
from the total excess of water in the outgoing air current (the sum of 
the amounts in columns 14 and 15 of Table 13). Thus the correct 
figure for the determination of the water given off by the combustion 
of the alcohol during the first period of the third test is 8.22 + 98.80 4- 
(31.68 -43.91), or 94.79 grams. It will be noticed that there was nearly 
twice as much water vapor in the chamber at the beginning of this 
test as at the end. The air at the beginning was comparatively moist, 
at the end unusually dry. The natural tendency of this condition 
would be to cause the evaporation, during the experiment, of some of 
the water that might have been condensed on the surface of the 
absorbers in the chamber. In experiments with man arrangements 
are made whereby the absorbers can be weighed, and it is not unusual 
to find a difference of over 100 grams in this weight during a six-hour 
period. It is quite probable that the excess of water found in this test 
was evaporated from the surface of the absorbers as the air became 
drier. 
In the first of the test experiments there was some irregularity in 
the first two periods, so that two totals are given, one for the whole 
experiment and one in which these first two periods are omitted. The 
difference in the results is not, however, appreciable. Including the 
first two periods, 99.9 per cent of the carbon dioxid and 100.5 per cent 
of the water theoretically given off by the combustion of the alcohol 
were measured, while if the first two periods are excluded these per- 
centages are 100 and 100.3, respectively. In the second test the 
agreement between the theoretical values and those measured was 
as close. The third test was not so satisfactory. More than the 
17951-No, 63 5 66 
theoretical amount of water was found in the air current, the amount 
being 101.1 per cent of that required. As was mentioned in the pre- 
ceding paragraph, the excess of water determined in this test is prob- 
ably due to the evaporation of some of that condensed on the absorbers 
at the beginning of the test, owing to the air becoming so much drier as 
the experiment progressed. 
A transcript of a page of the record book in which the calorimetric 
data are recorded by the operator at the " observer's table" during the 
progress of one of the electrical check experiments was given in Table 
1. A similar transcript of a page of the record made during the third 
of the alcohol check tests is given on page 67 (Table 14) as an illustra- 
tion of the method of observing and recording the fundamental data. 
The table contains the record for two consecutive hours, the six col- 
umns on the left side being given to the first and those on the right 
side to the second hour. 
The first three colums for each hour are similar to those in the record 
of the electrical tests. The fourth column for each hour gives the 
temperature of the air in the chamber, as determined by bridge No. 
5. The fifth column gives the difference of temperature of the ingoing 
and outgoing current of water by bridge No. 6. The sixth column 
gives occasional pairs of readings of the mercury thermometers, the 
difference serving as a check upon the bridge readings. As before, 
the mercury thermometers are taken as the standard, and the bridge 
readings are corrected so as to agree with the differences of the mer- 
cury thermometers. In the table the differences are one, three, and 
two hundredths of a degree, and the average of all these differences 
for the entire experiment is used as the correction of the temperature 
as measured by the bridge. This was 0.03°, and in Table 15, which 
gives a summary of the heat measurements for one period of the 
experiment, this correction is applied to the average of the bridge 
readings of the third column. The temperature of the copper wall is 
given by thermometer No. 7, as previously described on page 27, 
the reading being platted under " Remarks," as it is taken only occa- 
sionally. Under No. 8 are the readings of the mercury thermometer, 
whose bulb is in a pocket of the copper wall. That the readings of 
Nos. 5, 7, and 8 are not the same is due to different positions of the 
"zero" points. This introduces no error, as only the change in tem- 
perature is desired, and this is given alike by all three. 67 
Table 14.-Sample page from record book used for recording heat measurements, alcohol 
test experiment No. 3, May 27,1897. 
2 00 
02 
04 
06 
08 
10 
12 
14 
16 
18 
20 
22 
24 
26 
28 
30 
32 
34 
36 
38 
40 
42 
44 
46 
48 
50 
52 
54 
56 
58 
p 
k 
Time. 
o 
ND 
U 
4 
0 
4 
4 
4 
4 
1 
14 
14 
24 
24 
24 
14 
14 
1 
0 
4 
1 
14 
14 
14 
2 
24 
24 
2 
14 
14 
14 
14 
No. 1. 
Inner walls. 
© 
O 
toH toH top* top- 1-' >-1 top- top- top* top- toP- 
o o © o o 
top- top- top- top- ►-* top- i-' 1-' i-* top* top* top- bD 
No. 4. 
Moving air. 
21.18 
21.19 
21.19 
21.19 
21.19 
21.19 
21.19 
21.19 
21.19 
21.19 
21.17 
21.16 
21.15 
21.14 
21.13 
21.13 
21.13 
21.12 
21.11 
21. 11 
21.11 
21.10 
21.10 
21.10 
21.11 
21.11 
21.10 
21.11 
21.11 
21.11 
o 
Ci 
3 
p 
pt 
Inside temper- 
ature. 
3. 03 
3.04 
3.07 
3.09 
3.17 
3.20 
3.24 
3. 22 
3.24 
3. 35 
3. 49 
3.50 
3.50 
3. 50 
3. 48 
3.39 
3.35 
3.35 
3.35 
3. 32 
3. 32 
3.31 
3.31 
3.31 
3.24 
3.23 
3. 20 
3.19 
3.17 
3.17 
o 
Ci 
No. 6. 
Water by 
bridge. 
(14. 39 
11.08 
| 3.31 
(+.01 
o 
Ci 
£ 
Water ther- 
mometers. 
21.56; 22.38 
(10 kilo- 
<( grams at 
12: 20:35 
(10 kilo- 
< grams at 
(2:51:20 
No. 7,No.8. 
Remarks. 
3 00 
02 
04 
06 
08 
10 
12 
14 
16 
18 
20 
22 
24 
26 
28 
30 
32 
34 
36 
38 
40 
42 
44 
46 
48 
50 
52 
54 
56 
58 
h. m. 
A.M. 
Time. 
+ 
o 
4 
4 
0 
0 
4 
4 
4 
4 
4 
4 
4 
4 
4 
4 
4 
4 
4 
4 
4 
4 
4 
4 
14 
2 
14 
4 
4 
1 
1 
1 
1 
w 
2 
Inner walls. 
© 
f- 
top- mH : top* lip top »P P-* top- top* top- to toP- 
o o o o o o 
top- top- top- top- top- top- tor- top- top- t-• ton 
+ 
No. 4. 
Moving air. 
21.11 
21.12 
21.12 
21.13 
21.13 
21.13 
21.13 
21.14 
21.14 
21.14 
21.14 
21.15 
21.15 
21.15 
21.15 
21.15 
21.15 
21.16 
21.16 
21.16 
21.16 
21.16 
21.16 
21.16 
21.16 
21.16 
21.17 
21.17 
21.17 
21.17 
Ci 
No. 5. 
Inside temper- 
ature. 
3.17 
3.16 
3.16 
3.18 
3.19 
3.19 
3.19 
3.19 
3.17 
3.18 
3.18 
3. 20 
3.19 
3.18 
3.18 
3.18 
3.20 
3.22 
3.23 
3.26 
3.27 
3. 28 
3. 30 
3.28 
3. 27 
3.29 
3. 27 
3.28 
3. 28 
3.38 
o 
Ci 
No. 6. 
Water by 
bridge. 
(14.33 
|11.17 
3.16 
( + .03 
14.40 
11.14 
3.26 
+ .02 
o 
Q 
T' T 
zX iX 
W ater t h e r - 
mometers. 
21.50; 22.35 
(10 k i 1 o - 
< grams at 
( 3:22:32 
21.54; 22.38 
(10 k i 1 o - 
/ grams at 
( 3:53:00 
No. 7, No. 8. 
Remarks. 
The method of calculating the amount of heat carried out of the 
apparatus by the water is illustrated by Table 15, which summarizes 
the results for the fifth period of the third alcohol test experiment. 
The number of minutes occupied in the escape of 10 kilograms of water 
is shown by the figures in the first column. The mean difference 
between the temperatures of the incoming and outgoing water current 
during each of these intervals is shown in the third column. These 
temperatures are, however, as measured by the bridge. Applying the 
correction of three-hundredths of a degree obtained from the compari- 
son of the mercury and the copper thermometers above referred to, the 
corrected figures in the fourth column are obtained. These differences, 68 
multiplied by the number of kilograms of water, show., in the fifth col- 
umn, the number of calories of heat taken up by the water in its pas- 
sage to the chamber-that is, the heat measured by the water current. 
These values are taken in terms of calories at the mean temperature of 
the water while it was in the calorimeter. 
The figures in the last column show the temperature of the interior 
of the apparatus. It appears that this temperature was 21.12° at both 
the beginning and the end of the period; that is to say, the amount 
of heat in the metal walls and contiguous parts was the same at the 
beginning and end of the periods. There is, therefore, no correction 
to be applied for the heat capacity of the apparatus in this particular 
period. 
Table 15.-Summary for one period of alcohol test experiment No. 
l 3, May 37, 1897. 
(a) 
Time. 
Calorimetric measurements. 
Temper- 
ature of 
the inside 
by the 
copper 
ther- 
mometer 
No. 5. 
(b) 
Kilo- 
grams 
of 
water. 
(c) 
Differ- 
ence in 
tern pera- 
ture by 
bridge 
No. 6. 
(d) 
Corrected 
difference 
of tem- 
perature, 
c-.03. 
(e) 
Heat, 
b x d. 
h. 
m. 
3. 
°C. 
°C. 
Cals. 
°C. 
12 
41 
20 
21.12 
12 
47 
45 
2 
3.82 
3.79 
7. 58 
21.14 
1 
18 
49 
10 
3.63 
3 60 
36.00 
21.14 
1 
49 
45 
10 
2.86 
2.83 
28.30 
21.18 
2 
20 
35 
10 
3.14 
3. 11 
31.10 
21.17 
2 
51 
20 
10 
3.36 
3.33 
33.30 
21.11 
3 
22 
32 
10 
3.18 
3.15 
31.50 
21.15 
3 
53 
00 
10 
3.24 
3.21 
32.10 
21.17 
4 
24 
10 
10 
3.38 
3.35 
33.50 
21.11 
4 
54 
45 
10 
3.27 
3.24 
32.40 
21.11 
5 
26 
03 
10 
3.26 
3.23 
32. 30 
21.14 
5 
57 
00 
10 
3.35 
3. 32 
33.20 
21.14 
6 
28 
06 
10 
3.37 
3. 34 
33. 40 
21.11 
6 
42 
17 
4.6 
3.22 
3.19 
14.67 
379.35 
21.12 
In Tables 16, 17, and 18 are given summaries of the heat measure- 
ments, by six hour periods, in three of the alcohol test experiments. 
The number of grams of water vaporized, as recorded in the eleventh 
column, is multiplied by .592 to give the ''equivalent heat of water 
vaporized" in the twelfth column. The latter, added to "corrected 
heat measured," gives the "total heat" found. The number of grams 
of alcohol burned in each period, recorded in the third column, multi- 
plied by 6.415, the number of calories produced by the combustion of 
1 gram of the alcohol used, gives the number of calories that should 
have been produced in the alcohol flame. These are recorded in the 
fourteenth column as "calculated heat from alcohol burned." 69 
1897. 
Approximate period. 
Alcohol 
burned. 
Heat 
measured 
in terms 
of 
Ct to t2. 
Bange of 
tempera- 
ture of 
water 
t to t2. 
Heat 
measured 
reduced 
to C20. 
Change 
of tern - 
perature 
of calo- 
rimeter. 
Capacity 
correc- 
tion. 
Heatre- 
W ater 
vapor- 
ized. 
Equiva- 
lentheat 
of water 
vapor- 
ized. 
Total 
heat 
found. 
Calcu- 
lated heat 
from 
alcohol 
burned. 
Per cent 
meas- 
ured. 
moved 
by 
chim- 
ney. 
heat 
meas- 
ured. 
Gms. 
Cals. 
°C. 
Cals. 
°C. 
Cals. 
Cals. 
Cals. 
Gms. 
Cals. 
Cals. 
Cals. 
Apr. 27 
9 
a. m. to 
4 p.m 
112.2 
633.7 
3. 2-13.7 
636.1 
-0.03 
-1.8 
2 
636.3 
133.3 
78.9 
715.2 
719.8 
27 
4 
p.m. to 10 p. m..!... 
104.3 
581.6 
3. 7-13. 3 
583.7 
+0.15 
4-9.0 
2 
594.7 
123.4 
73.0 
667.7 
669.1 
27 
10 
p.m. to 
4 a. m 
103. 9 
590.6 
3.9-13. 3 
592.8 
-0.04 
-2.4 
2 
592.4 
119.7 
70.9 
663.3 
666.5 
28 
4 
a. m. to 11 a. m 
119.3 
669.2 
3.8-13. 3 
671.7 
+ 0.03 
+1.8 
2 
675.5 
137.6 
81.5 
757.0 
765. 3 
28 
11 
a. m. to 
3 p.m 
71.0 
402.1 
3.4-13.9 
403.6 
-0.01 
-0.6 
2 
405.0 
84.1 
49.8 
454.8 
455.5 
28 
3 
p.m. to 
9 p. m 
119.2 
670.2 
2.7-14.1 
• 672.8 
-0. 06 
-3.6 
2 
671.2 
130.4 
77.2 
748.4 
764.7 
28 
9 
p.m. to 
.3 a. m 
117.8 
663. 0 
2.9-14.0 
665.5 
-0.01 
-0.6 
2 
666.9 
138.5 
82.0 
748.9 
755.7 
29 
3 
a. m. to 
9 a. m 
111.8 
628.3 
3. 0-13. 6 
630.7 
0 
0 
2 
632.7 
132.4 
78.4 
711.1 
717.2 
29 
9 
a. m. to 
2 p.m 
95.9 
541.2 
3.1-13.5 
543. 4 
0 
0 
2 
545.4 
110.3 
65.3 
610.7 
615.2 
955.4 
5, 379. 9 
5,400. 3 
4-0.03 
4-1.8 
18 
5,420.1 
1,109. 7 
657.0 
6,077.1 
6,129.0 
99.15 
1897. 
Approximate period. 
Alcohol 
burned. 
Heat 
measured 
in terms 
of 
Ct tO t2- 
Range 
of tempera- 
ture of 
water 
t to t2. 
Heat 
measured 
reduced 
to C20• 
Change 
of tem- 
perature 
of calo- 
rimeter. 
Capacity 
correc-' 
tion. 
Heat 
removed 
by chim- 
ney. 
Corrected 
heat 
meas- 
ured. 
Water 
vapor- 
ized. 
Equiva- 
lent heat 
of water 
vapor- 
ized. 
Total 
heat 
found. 
Calcu- 
lated heat 
from 
alcohol 
burned. 
Per cent 
meas- 
ured. 
May 10 
10 
10 
11 
11 
6 a. m. to 12 m 
12 m. to 6 p. m •- 
6 p. m. to 12 p. m 
12'p. m. to 6 a. m 
6 a. m. to 12 m 
Gms. 
157.9 
147.7 
160.8 
165.8 
166.6 
Cals. 
888.4 
836.8 
936.2 
964.0 
964.3 
°C. 
3.2-13.3 
3. 7-13.4 
3.7-13.7 
3. 8-13. 6 
3.5-13.4 
Cals. 
891.7 
840.0 
939.7 
967.6 
968.0 
°C. 
0 
+ 0.05 
-0.02 
0 
-0.03 
Cals. 
0 
+3.0 
-1.2 
0 
-1.8 
Cals. 
2.5 
2.5 
2.5 
2.5 
2.5 
Cals. 
894.2 
845.5 
941.0 
970.1 
968.7 
Gms. 
185.6 
178.3 
182.6 
189.5 
189.0 
Cals. 
109.8 
105.6 
108.1 
112.2 
111.9 
Cals. 
1,004.0 
951.1 
1, 049.1 
1, 082.3 
1,080.6 
Cals. 
1,012.9 
947.5 
1,031.5 
1, 063. 6 
1,068.7 
798.8 
4, 589. 7 
4,607.0 
0 
0 
12.5 
4, 619. 5 
925.0 
547.6 j 5,167.1 
5,124.2 100. 84 
Table 16.-Summary of the calorimetric measurements in alcohol test experiment No. 1, April 27-^9, 1897. 
Table 17.-Summary of the calorimetric measurements in alcohol test experiment No. 2, May 10-11, 1897. 70 
1897. 
Approximate period. 
Alcohol 
burned. 
Heat 
measured 
in terms 
of 
Ct to t2* 
Range 
of tempera- 
ture of 
water 
t to t^. 
Heat 
measured 
reduced 
to Cgo* 
Change 
of tem- 
perature 
of calo- 
rimeter. 
Capacity 
correc- 
tion. 
Heat 
removed 
by chim- 
ney. 
Corrected 
heat 
meas- 
ured. 
Water 
vapor- 
ized. 
Equiva- 
lent heat 
of water 
vapor- 
ized. 
Total 
heat 
found. 
Calcu- 
lated heat 
from 
alcohol 
burned. 
Per cent 
meas- 
ured. 
Gms. 
Cals. 
°C. 
Cals. 
°C. 
Cals. 
Cals. 
Cals. 
Gms. 
Cals. 
Cals 
Cals. 
May 26 
1 
a. m. to 8 
a. in 
72.5 
402.8 
15.5-17.5 
403.0 
-0.15 
- 9.0 
1.5 
395.5 
94.8 
56.1 
451. 6 
465. 0 
26 
8 
a. m. to 1 
p. m 
94.6 
537.6 
4. 3- 8.6 
540.1 
-0.20 
-12.0 
2.5 
530.6 
104.6 
61.9 
592. 5 
606.9 
26 
1 
p. m. to 6 
p. m 
108.0 
626. 1 
7. 0-12. 0 
628.1 
+ 0.15 
+ 9.0 
2 
639.1 
127.9 
75.7 
714.8 
692. 8 
26 
6 
p. m. to 1 
a. m 
91.3 
493.0 
11.1-15. 2 
493.8 
+0.12 
+ 7.2 
2 
503.0 
119.9 
71.0 
574.0 
585. 7 
27 
1 
a. m. to 7 
a. in 
68.4 
379.4 
11.2-14.4 
380.0 
0 
0 
2 
382.0 
90.7 
53.7 
435.7 
438.8 
27 
7 
a. m. to 1 
p. m 
70.6 
393.1 
11.0-14.5 
393.8 
0 
0 
2 
395.8 
90.0 
53.3 
449.1 
452. 9 
505.4 
2,832.0 
2,838. 8 
- .08 
- 4.8 
12.0 
2,846.0 
627.9 
371.7 
3, 217.7 
3. 242.1 
99. 25 
Table 18.-Summary of the calorimetric measurements in alcohol test experiment No. 3, May 26-27, 1897. 71 
In Table 19 are summarized the principal results of all the alcohol 
check tests which were made up to the ninth, except the preliminary 
tests, in which the methods of manipulation were being worked out, 
and two tests the completion of which was prevented by accident. 
These individual experiments continued from 5 to 78 hours each, the 
total time being 317 hours. The rate of burning of the alcohol ranged 
from 10 to 27 grams per hour, and the strength of the alcohol from 
90.26 to 90.77 per cent. Most of the tests were made in alternation 
with experiments with men, the object being to test the accuracy of 
the apparatus before and after each of the latter experiments. 72 
No. 
Date. 
Duration. 
Alcohol 
burned. 
Carbon dioxid. 
Water. 
Heat. 
Required. 
Found. 
Ratio of 
amount 
found to 
amount 
required. 
Required. 
Found. 
Ratio of 
amount 
found to 
amount 
required. 
Required. 
Found. 
Ratio of 
amount 
found to 
amount 
required. 
1 
1897. 
April 27-29 
h. m. 
52 31 
Grams. 
955.4 
Grams. 
1, 657.2 
Grams. 
1, 657. 6 
Per cent. 
100.0 
Grams. 
1,106.1 
Grams. 
1,109.7 
Per cent. 
100.3 
Calories. 
6.129.0 
Calories. 
6, 077.1 
Per cent. 
99.15 
2 
May 10-11 
29 56 
798.8 
1,385.6 
1, 384. 4 
99.9 
924.8 
925.0 
100.0 
5,124.2 
5,167.1 
100. 84 
3 
May 26-27 
33 50 
505.4 
876. 7 
887.8 
101.3 
585.1 
627.9 
[107.3] 
3,242.1 
3,217. 7 
99.25 
4 
October 27-28 
34 33 
797.7 
1, 384. 8 
1,335.7 
[96. 5] 
925.7 
1.007.9 
[108.8] 
5,120. 5 
5,141.5 
100.41 
5 
November 2-3 
35 09 
788.2 
1,365.1 
1, 376. 7 
100.8 
912.3 
920.8 
100.9 
5, 048. 4 
5,050.0 
100.03 
6 
December 2 
11 39 
245.3 
423.1 
417.6 
98.6 
283.7 
287.5 
101.3 
1, 564. 8 
1,556.8 
99.48 
7 
1898. 
January 6 
5 50 
112.2 
193.5 
193.5 
100.0 
129.8 
131.3 
101.2 
715.7 
731.1 
102.15 
8 
January 24-27 
77 57 
1, 607.8 
2, 784. 4 
2, 769. 7 
99.5 
1,860.8 
1, 881. 6 
101.1 
10,294.7 
10, 268. 5 
99.74 
9 
May 9 
3d 55 
699.7 
1,206. 9 
1,198.9 
99.4 
809.3 
807.9 
99.8 
4, 463. 4 
4, 466.0 
100.05 
Total a 
317 20 
9. 892. 5 
9, 886. 2 
99.9 
6, 026. 8 
6, 063.8 
100.6 
41,702. 8 
41,675.8 
99.93 
a Omitting the carbon dioxid and water in test No. 4 and the water in test No. 3. 
Table 19.-Summary of nine alcohol test experiments. 73 
It will be noticed that in No. 4 the determination of carbon dioxid 
was unsatisfactory, though we were unable to decide whether the error 
was due to imperfect sampling or other cause. The result is not in- 
cluded in the average of Table 19. The measurement of the heat was, 
however, very close to the theoretical value. Almost immediately after 
the close of this test another was made, in which all the results were 
very closely in accord with the theoretical values. If the determi- 
nations of carbon dioxid in test No. 4 are omitted, the maximum vari- 
ation of the amounts determined from the theoretical amounts given 
off by the combustion of the alcohol was 1.4 per cent and the average 
variation only 0.1 per cent, or 1 part in 1,000 from the theoretical. 
In tests Nos. 3 and 4 the determinations of water were unsatisfac- 
tory. We are inclined to attribute the errors to variations in the 
amounts of water condensed upon the absorbers. These results, like 
that for carbon dioxid in No. 4, are omitted from the averages. Omit- 
ting the determinations of water in tests Nos. 3 and 4, the maximum 
variation as actually determined from the theoretical amount was 1.2 
per cent and the average variation only 0.6 per cent. In test No. 7 the 
proportion of heat measured was larger than usual. It will be observed, 
however, that this test continued only through one period of six hours. 
Some time is required to get the apparatus in temperature equilibrium, 
and the heat measurements of the first experimental period are fre- 
quently incorrect on this account. The omission of this experiment 
would not materially affect the total averages. Omitting this expert 
ment, the maximum variation of the heat actually measured from the 
theoretical amount was 0.8 per cent and the average variation only 0.1 
per cent. 
SUMMARY OF TEST EXPERIMENTS. 
The accuracy of the methods for the determination of carbon dioxid, 
water, and heat was tested by heat generated in the chamber by pass- 
ing an electric current through a resistance coil and by burning ethyl 
alcohol within the chamber. In the electrical tests the measurements 
of heat generated and found were practically identical, the differences 
between the theoretical and actual results averaging about 0.1 per 
cent-that is, about 1 part in 1,000. In the alcohol tests the average 
amounts found by actual experiment were: For carbon, 99.9 per cent; 
hydrogen, 100.6 per cent; and heat, 99.9 per cent, respectively, of the 
theoretical amounts. 
The determinations of carbon dioxid and water made by burning 
large quantities of alcohol in the respiration chamber agree reasonably 
well with each other and with the theoretical amounts. The variations, 
indeed, are not greater than are found in ordinary laboratory experi- 
ence when alcohol is burned in the combustion furnace by the usual 
methods of organic analysis. 
The agreement between the results given by the respiration calorim- 
eter and the bomb calorimeter for the heat of combustion of alcohol, as 
shown in Table 35, is also very satisfactory when the great difference 
in the circumstances of the experiments is taken into consideration. 
In the bomb calorimeter a fraction of a gram of alcohol is absorbed in a 74 
small block of cellulose or inclosed in a gelatin capsule and placed in a 
steel cylinder of perhaps half a liter capacity, which is filled with oxygen 
at a pressure of 20 atmospheres. An electric current passes through a 
fine iron wire, melts the latter, and ignites the alcohol, which is com- 
pletely oxidized in an instant. The increase of temperature of the 
bomb and the water in which it is immersed gives the quantity of heat 
evolved, and from that the heat of combustion of a gram of alcohol 
is computed. In the respiration calorimeter, on the other hand, the 
alcohol contained in a small lamp burns quietly for many hours in a 
chamber ten thousand times as large as the bomb. Oxygen is supplied 
by a continuous current of air pumped through the apparatus, and a 
considerable portion of the heat of combustion is carried away in the 
latent heat of the water vapor produced. The sum of the heat meas- 
ured by the calorimeter and the latent heat of the water vapor collected 
gives the total heat produced by the burning alcohol. That the average 
of a series of nine experiments should vary less than 0.1 per cent from 
the average of the determinations with the bomb calorimeter seems 
a gratifying result. Taken in connection with the electrical tests and 
the determinations of carbon dioxid and water already given, the results 
show that the respiration calorimeter is an instrument of precision and 
abundantly capable of doing the work for which it was designed. 
EXPERIMENTS WITH A MAN. 
The apparatus previously described was designed exclusively for 
experiments with human subjects. There were several reasons for 
beginning with men rather than with domestic animals. The study of 
human nutrition is very important. In the earlier development of the 
work, when many difficulties were to be overcome, it was very desira- 
ble to have inside the apparatus an intelligent person who could make 
and record important observations during the experiment, rather than 
an animal whose movements, even, could not be controlled. Indeed, the 
most advantageous way to develop methods and apparatus for experi- 
ments with animals is through such preliminary experience with men. 
The results of the experience thus far gained are now being utilized 
in planning apparatus and methods to be used not only with small 
animals, as rabbits, sheep, and dogs, but also with larger animals, as 
horses, oxen, and cows. 
PLAN OF THE EXPERIMENTS 
The general plan of the experiments here described is the same as 
that of the respiration experiments previously reported.1 For each 
experiment a diet is selected which furnishes the different nutrients 
in the amounts and proportions appropriate to the question under 
investigation. This diet is followed for eight days, of which the last 
four constitute the period of the experiment proper. In the first or 
preliminary period, of four days, the analyses of feces and urine are 
made, the data thus sufficing for a digestion and nitrogen metabolism 
experiment. On the evening of the fourth day the subject enters the 
1 U. S. Dept. Agr., Office of Experiment Stations Bui. 44. 75 
respiration chamber, though the actual respiration calorimeter experi- 
ment does not begin until 7 o'clock on the morning of the fifth day. The 
night sojourn in the apparatus suffices to get the temperature of the 
apparatus and its content of carbonic acid and water into equilibrium, 
so that accurate measurements may begin with the morning of the fifth 
day and continue until 7 o'clock on the morning of the ninth day, thus 
making tlie duration of this experiment exactly four days. The deter 
minations of carbon dioxid, water vapor, and heat are made in six- 
hour periods, so that the complete data for an experiment show the 
total amounts of these compounds given off from the body during the 
periods ending at 1 p. m., 7 p. m., 1 a. m., and 7 a. m. of each day of 
the experiment. The urine is also collected and the nitrogen deter- 
mined for corresponding periods. 
A definite schedule is planned before the commencement of the ex- 
periment, which is followed with reasonable closeness by the subject 
in the chamber and by those having the experiment in charge. The 
accompanying schedule or program shows the daily routine followed in 
each of the experiments here reported: 
Table 20.-Daily program of the rest and work experiments here reported. 
Time. 
Rest experiment (No. 9). 
Work experiment (No. 6). 
7.00 a. m 
Rise 
Rise. 
Pass urine 
Pass urine. 
"Weigh self, stripped 
Weigh self, stripped and dressed. 
Weio-h absorbers 
Weigh absorbers. 
Collect drip. 
Breakfast. 
Collect drip 
7.45 a. m .. 
Breakfast 
Weigh self. 
8.20 a. in .... 
Begin work. 
10 minutes' rest. 
10.20 a. in 
10.30 a. in 
Drink 200 grams water 
Weigh self; drink 200 grams water. 
Stop work. 
Pass urine. 
12.30 p. in . . 
1.00 p. in . 
Pass urine 
Collect drip. 
Collect drip 
Weigh self. 
Weigh absorbers 
Weigh absorbers. 
1.15 p. m . 
Dinner 
Dinner. 
Drink 200 grams water. 
Weigh self. 
1.50 p. in 
Begin work. 
10 minutes' rest. 
3.50 p.m 
4.00 p. m 
Drink 200 grams water 
Weigh self; drink 200 grams water. 
Stop work; weigh self. 
Supper. 
6.00 p. in ........... 
6.30 p. in .... 
Supper 
Change underclothing. 
Weigh self, stripped and dressed. 
7 00 p.in .. -. -. - 
Pass urine. ... ......... ..... 
Pass urine. 
Collect drin 
Collect drip. 
Weigh absorbers. 
Weigh absorbers 
10.00 p.m .... 
Drink 200 grams water 
Weigh self. 
Drink 200 grams water. 
Weigh self, stripped 
Retire 
Retire. 
1.00 a. in 
Pass urine 
Pass urine. 76 
In the rest experiment the subject was as quiet as he well could be. 
In the four days of the preliminary period he moved about but little 
and engaged in no considerable amount of either muscular or mental 
labor. During the four days passed in the chamber he was likewise 
quiet. The only muscular work done was that involved in dressing, 
putting up and taking down the folding chair, table, and bed, weighing 
himself and the absorbers, taking his meals, and caring for the excreta. 
He passed a large part of the time in reading and sleeping. 
In the work experiment the subject was engaged in active muscular 
labor. The energy of the external muscular work done was entirely 
transformed into heat within the chamber. The larger part was first 
transformed into electrical energy by a small dynamo which was belted 
to the wheel of a stationary bicycle, and was then transformed into heat 
by an electric lamp through which the current passed. A small por- 
tion was transformed into heat by the friction of this bicycle dynamo 
or ergometer. The heat thus produced was measured with the heat 
given off from the body. The muscular work was continued for about 
eight hours per day, and the external muscular power as roughly 
measured by this ergometer was estimated to be equivalent to not far 
from 250 calories per day. This measurement was not as accurate as 
desirable and a special apparatus is now being constructed for the pur- 
pose. However, any error in measuring the work done does not affect 
the determination of the total energy since this is also measured by 
the calorimeter. It only affects the percentage found for the muscular 
work done-that is, the efficiency of the man as a machine. 
All the furniture and bedding were weighed at the beginning and 
at the end of each experiment, but there was no appreciable change 
in weight. In the work experiment there was an apparent gain in 
weight of 20 grams in a total of 21,000, a number so small as to be 
within the limits of error in weighing. 
In the analyses of the food, feces, and urine the methods employed 
were essentially those adopted by the Association of Official Agri- 
cultural Chemists,1. with such modifications as experience and cir- 
cumstances have shown desirable. The preparation, sampling, and 
analysis of the food materials was much the same as that described in 
a previous publication.2 
In the rest experiment (No. 9) and those made later, most of the 
food materials were divided into portions proper for individual meals 
and put into jars. In this way the sampling was rendered more 
accurate and the labor involved in the preparation of the meals during 
an experiment was reduced to a minimum. 
The treatment of the feces and urine was the same as that already 
described in the publication just referred to. 
1 U. S. Dept. Agr., Division of Chemistry Bui. 46. 
2U. S. Dept. Agr., Office of Experiment Stations Bui. 44. 77 
The heats of combustion of the food, feces, and urine were determined 
in the usual manner by means of the bomb calorimeter.1 
The determinations of carbon and hydrogen were made in the usual 
manner. The partially dried samples of food, feces, and urine were 
burned with cupric oxid with the aid of a current of oxygen, and the 
water absorbed by sulphuric acid, the carbon dioxid by potassium 
hydroxid. In some of the later experiments, as in No. 9, modifications 
of this method were employed, by which the time required for com- 
bustion was materially reduced. These modifications have not yet 
been described. 
The percentage composition and heats of combustion of the food used 
in the two experiments here described, and of the feces for these ex- 
periments are shown in Table 21. 
Table 21.-Composition of food materials, etc. 
Laboratory No. 
Food materials, etc. 
| Experiment No. 
Nitrogen. 
Carbon. 
Hydrogen. 
W ater. 
Protein (N.X 
6.25). 
Fat. 
Carbohydrates, 
Ash. 
Heats of com- 
bustion per 
gram, a 
P. ct. 
P. ct. 
P.ct. 
P. ct. 
P. ct. 
P. ct. 
P. ct. 
P.ct. 
Cals. 
2789 
Beef, fried 
6 
4. 77 
21.28 
3.05 
60.3 
29.8 
8.7 
... 
2.09 
2.421 
2835 
Beef, cooked 
9 
4.10 
16.35 
2. 25 
67.3 
25.6 
5.4 
... 
1.56 
1.928 
2788 
Ham, deviled 
6 
2.64 
35.67 
4.48 
42.2 
16.5 
36.9 
... 
4.03 
4.353 
2790 
Eggs 
6 
2.24 
14.39 
2.19 
73.2 
14.0 
11.3 
... 
.98 
1.928 
2793 
Butter 
6 
.16 
62.82 
10.34 
9.3 
1.0 
87.3 
... 
2. 39 
7.954 
2833 
Do 
9 
.19 
62. 68 
10.27 
10.2 
1.2 
84.8 
... 
3.80 
7.761 
2799 
Milk, whole 
6 
.48 
8.27 
1.23 
85.3 
3.0 
5.4 
5 
6 
.69 
.935 
2836 
Milk, skimmed 
9 
.52 
4. 04 
.57 
90.7 
3.3 
.1 
5 
2 
.•78 
.393 
2803 
Bread, white 
6 
1.33 
25.45 
3.85 
43.9 
8.3 
1.6 
45 
0 
1.24 
2.540 
2834 
Bread 
9 
1.34 
24.53 
3.54 
44.7 
8.4 
.2 
44 
3 
2.44 
2.400 
2830 
Breakfast food, wheat 
9 
1.58 
41.32 
5. 78 
7.5 
9.9 
1.6 
77 
7 
3. 26 
4.071 
2829 
Ginger snaps • 
9 
.96 
44.45 
6.48 
5.2 
6.0 
9.5 
75 
6 
3. 70 
4. 358 
2831 
Breakfast food, maize 
9 
1.78 
44.34 
6.45 
5.6 
11.1 
8.7 
71 
1 
3.46 
4.444 
2791 
Beans, baked 
6 
1.15 
12.44 
1.73 
71.4 
7.2 
.4 
19 
2 
1.86 
1. 222 
2792 
Pears, canned 
6 
.05 
7.01 
1.18 
81.4 
.3 
.2 
17 
9 
.24 
.759 
2786 
Sugar 
6 
42.10 
6.48 
100 
0 
3. 963 
2832 
Do 
9 
42. 10 
6.48 
100 
0 
3. 960 
2808 
Feces 
6 
1.29 
10.64 
1.56 
78.6 
8.1 
4.1 
5 
6 
3.64 
1.194 
2838 
Do 
9 
1.19 
12. 60 
1.74 
72.9 
7.4 
3.9 
11 
4 
4.40 
1.343 
a As determined. 
EXPLANATION OF THE TABLES. 
The results obtained in two experiments, Nos. 9 and 6 respectively, 
are shown in the following tables. Of these two experiments, No. 9 
was a "rest" experiment, and was made in January, 1898. No. 6 was 
a "work" experiment, in which the subject rode the stationary bicycle 
1 For detailed description of improved forms of bomb calorimeter and accessory 
apparatus, see Connecticut Storrs Sta. Rpt. 1897, pp. 199-211. 78 
about eight hours a day, and was carried on in May, 1897. Only such 
of the data are given here as it is believed will be sufficient for a clear 
understanding of the experiments. The order of the experiments is 
reversed to follow the more logical order of rest and work. 
The daily diet in each of these two experiments was determined before 
the commencement of the preliminary period, and was followed through- 
out the eight days of the experiment, of which, as previously explained, 
the last four days were spent in the respiration chamber. 
The daily menu in each experiment was as follows: 
Table 22.-Daily menu-rest experiment (No. 9). 
Breakfast. 
Grams. 
Dinner. 
Grams. 
Supper. 
Grams. 
Cooked beef 
100 
Cooked beef 
150 
Butter 
15 
Butter 
15 
Butter 
20 
Skimmed milk 
390 
Skimmed milk 
160 
Skimmed milk 
210 
Bread 
25 
Bread 
25 
Bread . 
50 
Breakfast food, wheat. 
75 
Breakfast food, maize.. 
50 
Breakfast food, wheat. 
50 
Ginger snaps 
60 
Sugar 
25 
Sugar 
>a 
Sugar 
30 
Coffee, about 
300 
Coffee, about 
300 
Coffee, about 
300 
Breakfast. 
Grains. 
Dinner. 
Grains. 
Supper. 
Grams. 
"Deviled "ham... 
20 
Cooked beef 
100 
30 
Boiled eggs 
Butter 
30 
25 
Rutter 
20 
Milk 
50 
Milk 
600 
Milk 
200 
Bread 
125 
175 
Bread 
150 
Baked beans 
125 
15 
Susar 
15 
Canned pears 
300 
300 
Coffee, about 
300 
Sugar 
20 
Coffee, about 
300 
Table 23.-Daily menu-work experiment (No. 6). 
Tables 24 and 25 show the amount of carbon dioxid eliminated in 
the rest and work experiments, respectively. The data reported show 
the number of liters of air in the ventilating air current and the milli- 
grams per liter of carbon dioxid in the incoming and in the outgoing 
air during each six-hour period of the experiment. From these data 
the total excess of carbon dioxid in the outgoing over that of the 
incoming air is calculated. The analysis of the residual air in the 
apparatus at the close of each period gives data for the calculation of 
the increase or decrease of the carbon dioxid in the chamber at the 
end as compared with the beginning of each period. Applying this 
correction as explained on page 65, the actual elimination of carbon 
dioxid by the subject is obtained for each six-hour period. From this 
is calculated the total weight of carbon exhaled. 79 
Table 24.-Record of carbon dioxid in rest experiment (No. 9). 
Day. 
Period. 
Ventila- 
tion : 
number 
of liters 
of air. 
Carbon dioxid per liter. 
Total 
excess 
in out- 
going 
air. 
Correc- 
tion for 
carbon 
dioxid 
in appa- 
ratus. 
Cor- 
rected 
weight 
carbon 
dioxid 
exhaled 
by sub- 
ject. 
Total 
weight 
carbon 
exhaled 
in 
carbon 
dioxid. 
In in- 
coming 
air. 
In out- 
going 
air. 
Excess 
inout- 
going 
air. 
Liters. 
Mg. 
Mgs. 
Mgs. 
Grams. 
Grams. 
Grams. 
Grams. 
1 
7 a. m. to 1 p. m 
25, 712 
0. 580 
9.556 
8.976 
230.8 
+ 3.1 
233.9 
63.8 
1 p. in. to 7 p. in 
25, 987 
.563 
9. 686 
9.123 
237.1 
+ 14.9 
252.0 
68.7 
7 p. m. to 1 a. m 
26, 785 
. 622 
8.997 
8. 375 
224. 3 
14. 7 
209. 6 
57. 2 
1 a. m. to 7 a. m 
26, 065 
.593 
5.891 
5. 298 
138.1 
- 2.5 
135.6 
37.0 
Total 
104, 549 
830.3 
4- . 8 
831.1 
226.7 
Q 
7 a. in. to 1 p. in 
27,057 
.571 
8.750 
8.179 
221. 3 
+11.3 
232. 6 
63.4 
1 p. m. to 7 p. in 
25, 878 
.560 
9.109 
8.549 
221.3 
+ 5.3 
226. 6 
61.8 
7 p. m. to 1 a. m 
26,652 
.612 
9.556 
8. 944 
238.4 
12. 6 
225. 8 
61. 6 
la. m. to 7 a. m 
26, 011 
.629 
5. 765 
5.136 
133.6 
- 5.2 
128.4 
35.0 
Total 
105, 598 
814.6 
- 1.2 
813.4 
221.8 
3 
7 a. in. to 1 p. m 
26, 342 
.624 
9.139 
8.515 
224. 3 
+19. 2 
243. 5 
66. 4 
1 p. m. to 7 p. in 
26, 492 
.721 
9.349 
8. 628 
228.6 
+ -3 
228. 9 
62.4 
7 p. m. tol a. m 
26,147 
.722 
9. 326 
8. 604 
225.0 
10. 6 
214 4 
58 5 
la. m. to 7 a. m 
25,163 
.727 
5.917 
5.190 
130.6 
- 8.1 
122. 5 
33.4 
Total 
104,144 
808.5 
+ -8 
809.3 
220.7 
4 
7 a. in. to 1 p. in 
26, 427 
.710 
8.973 
8. 263 
218. 4 
+21.2 
239 6 
65 3 
1 p. in. to 7 p. m 
25, 731 
.735 
9.525 
8. 790 
226.2 
+ . 3 
226 5 
61 8 
7 p. in. to 1 a. m 
26, 046 
.636 
10.021 
9.385 
244. 5 
14. 0 
230. 5 
62. 9 
la. in. to 7 a. m 
26, 338 
.612 
5.870 
5. 258 
138.5 
- 9.1 
129.4 
35.3 
Total 
104, 542 
827. 6 
1.6 
826. 0 
225. 3 
Total, four days 
418, 833 
3, 279.8 
894. 5 
Average per day... 
104,708 
- 
819.9 
223.6 
Day. 
Period. 
Ventila- 
tion; 
number 
of liters 
of air. 
Carbon dioxid per liter. 
Total 
excess 
in out- 
going 
air. 
Correc- 
tion for 
carbon 
dioxid 
in appa- 
ratus. 
Cor- 
rected 
weight 
carbon 
dioxid 
exhaled 
by sub 
ject. 
Total 
weight 
carbon 
exhaled 
in 
carbon 
dioxid. 
In in- 
coming 
air. 
In out- 
going 
air. 
Excess 
in out- 
going 
air. 
Liters. 
Mg. 
Mgs. 
Mgs. 
Grains. 
Grams. 
Grams. 
Grams. 
1 
7 a. m. to 1 p. m 
25,329 
0.905 
17.190 
16.285 
412.5 
4-69.9 
482.4 
131.57 
1 p. m. to 7 p. in 
21,497 
. 994 
23.520 
22. 526 
484.2 
4-11 4 
495 6 
135 17 
7 p. m. to 1 a. m 
22,274 
. 771 
13. 769 
12. 998 
289 5 
52 2 
237 3 
64 72 
1 a. m. to 7 a. m 
22,173 
.730 
7.435 
6.705 
148.7 
-29.6 
119.1 
32.50 
Total 
91, 273 
1, 334. 9 
- .5 
1, 334. 4 
363.96 
2 
7 a. in. to 1 p. m 
23, 364 
. 710 
15.263 
14.553 
340 0 
4 76.4 
416. 4 
113 55 
1 p. in. to 7 p. in 
23, 356 
. 647 
20. 613 
19. 966 
466. 3 
3. 8 
462. 5 
126.17 
7 p. m. to 1 a. m 
23,999 
. 587 
13.651 
13.064 
313 5 
60 6 
252 9 
68 97 
1 a. m. to 7 a. m 
23, 541 
.606 
6. 381 
5.775 
136.0 
-13.4 
122.6 
33. 40 
Total 
94, 260 
1, 255.8 
- 1.4 
1, 254.4 
342. 09 
3 
7 a. in. to I p. m 
23,147 
. 516 
16.143 
15.627 
361. 7 
4-90. 5 
452.2 
123 35 
1 p. in. to 7 p. m 
23, 044 
. 612 
21 070 
20 458 
471. 4 
33 8 
437 6 
119 35 
7 p. in. fol a. in 
22,423 
. 912 
13.474 
12. 562 
281. 6 
41.1 
240.5 
65 60 
1 a. m. to 7 a. m 
23, 344 
.678 
6. 925 
6. 247 
145.8 
-10.4 
135.4 
36.94 
Total 
91, 958 
1,260.5 
4-5.2 
1, 265. 7 
345.24 
Table 25.-Record of carbon dioxid in work experiment (No. 6). 80 
Table 25.-Record of carbon dioxid in work experiment (No. 6)-Continued. 
Day. 
Period. 
Ventila- 
tion ; 
number 
of liters 
of air. 
Carbon 
dioxid per liter. 
Total 
excess 
in out- 
going 
air. 
Correc- 
tion for 
carbon 
dioxid 
in appa- 
ratus. 
Cor- 
rected 
weight 
carbon 
dioxid 
exhaled 
by sub- 
ject. 
Total 
weight 
carbon 
exhaled 
in 
carbon 
dioxid. 
In in- 
coming 
air. 
In out- 
going 
air. 
Excess 
in out- 
going 
air. 
Liters. 
Mg. 
Mgs. 
Mgs. 
Grains. 
Grams. 
Grams. 
Grams. 
4 
7 a. m. to 1 p. m 
23, 268 
.523 
15.612 
15.089 
351.1 
+ 74.6 
425.7 
116. 08 
1 p. m. to 7 p. m 
23, 031 
.574 
19. 583 
19. 009 
437.8 
-32.5 
405.3 
110. 55 
7 p. m. to 1 a. m 
23, 833 
.568 
12.270 
11.702 
278.9 
-50.0 
228.9 
62.41 
1 a. m. to 7 a. m 
25,757 
.999 
6. 719 
5.720 
147.3 
+ .8 
148.1 
40.40 
Total 
95,889 
1, 215.1 
- 7.1 
1,208. 0 
329. 44 
Total, four days... 
373, 380 
5, 062.5 
1,380.73 
Average per day... 
93, 345 
1, 265.6 
345.18 
Tables 26 and 27 show the amount of water exhaled in the rest 
and work experiments, respectively. The data include the number of 
liters of air in the ventilating air current and the milligrams per liter 
of water vapor in the incoming air and in the outgoing air after pass- 
ing the freezers, which latter condense the major portion of the water 
vapor in the outgoing air current. From these data, together with the 
amount of water condensed in the freezers, the amount condensed in 
the chamber as "drip," and the determinations of residual water vapor 
in the chamber, is computed the total water exhaled by the subject. 
Table 26.-Record of water in rest experiment (No. 9). 
Day. 
Period. 
Ventila- 
tion ; 
number 
of liters 
of air. 
Water per liter. 
Total 
excess 
in out- 
going 
air. 
Con- 
densed 
in 
freez- 
ers. 
Con- 
densed 
in 
cham- 
ber. 
Correc- 
tion for 
water 
vapor 
in 
cham- 
ber. 
Total 
water 
ex- 
haled. 
In in- 
coming 
air. 
In out- 
going 
air. 
Excess 
in out- 
going 
air. 
Liters. 
Mg. 
Mgs. 
Mg. 
Grams. 
Grams. 
Grams. 
Grams. 
Grams. 
1 
7 a. m. to 1p.m.. 
25, 712 
0.895 
1. 301 
0. 406 
10.4 
207.5 
14.0 
-13.5 
218.4 
1 p.m. to 7 p. m.. 
25,987 
.816 
1.258 
.442 
11.5 
230.2 
22.0 
+15.3 
279.0 
7 p. m. to 1 a. m.. 
26, 785 
.792 
1.184 
.392 
10.5 
238.1 
-18.0 
+ 3.1 
233.7 
1 a. m. to 7 a. m.. 
26, 065 
.731 
1.143 
.412 
10.7 
223.0 
- 6.9 
226.8 
Total 
104,549 
43.1 
898.8 
18.0 
- 2.0 
957.9 
2 
7 a. m. to 1 p. m.. 
27,057 
.737 
1.187 
.450 
12.2 
212.0 
3.0 
+ .5 
227.7 
1 p. m. to 7 p. m.. 
25,878 
.821 
1.232 
.411 
10.6 
213.7 
- 1.0 
+ 2.2 
225.5 
7 p. m. to 1 a. m.. 
26, 652 
.749 
1. 221 
.472 
12.6 
230.7 
-21.0 
+ 4.6 
226.9 
1 a. m. to 7 a.m.. 
26, 011 
.723 
1.219 
.496 
12.9 
208.4 
- 9.7 
211.6 
Total 
105,598 
48.3 
864.8 
-19.0 
- 2.4 
891.7 
3 
7 a. m. to 1 p. m.. 
26, 342 
.764 
1.320 
.556 
14.6 
202.1 
13.1 
+ 3.8 
233.6 
1 p. m. to 7 p. m.. 
26, 492 
. 772 
1.206 
.434 
11.5 
208.3 
4.0 
+ -5 
224.3 
7 p. m. to 1 a. m.. 
26,147 
.740 
1. 088 
.348 
9.1 
230.6 
- 8.0 
+10.5 
242.2 
1 a. m. to 7 a. m.. 
25,163 
.799 
1.271 
.472 
11.9 
209.0 
-15.1 
205.8 
Total 
104,144 
47.1 
850.0 
9.1 
- .3 
905.9 
4 
7 a. m. to 1p.m.. 
26, 427 
.793 
1.293 
.500 
13.2 
209.2 
41.8 
+ 3.9 
268.1 
1 p. m. to 7 p. m.. 
25, 731 
.726 
1.273 
.547 
14.1 
206.2 
37.8 
+ 2.3 
260.4 
7 p. m. to 1 a. m.. 
26, 046 
.776 
1.128 
.352 
9.2 
235.5 
- 6.7 
+ 5.4 
243. 4 
1 a. m. to 7 a. m.. 
26, 338 
.715 
1.109 
.394 
10.4 
234.6 
4.2 
-11.6 
237.6 
Total 
104,542 
46.9 
885. 5 
77.1 
1, 009.5 
Total, four days 
418.833 
185. 4 
3,499.1 
85.2 
4. 7 
3,765.0 
Avg. per day... 
104. 708 
941. 3 81 
Day. 
Period. 
Ventila- 
tion ; 
num- 
ber of 
liters of 
air. 
Water per liter. 
Total 
excess 
in out- 
going 
air. 
Con- 
densed 
in freez- 
ers. 
Con- 
densed 
in cham- 
ber. 
Correc- 
tion for 
water 
vapor 
in cham- 
ber. 
Total 
water 
exhaled. 
In in- 
coming 
air. 
In out- 
going 
air. 
Excess 
in out- 
going 
air. 
Liters. 
Mgs. 
Mgs. 
Mgs. 
Grams. 
Grams. 
Grams. 
Grams. 
Grams. 
1 
7 
a. m. to 1 p. m 
25,329 
1.164 
1 
369 
0.205 
5.2 
279.1 
483.2 
4-3.7 
1 
p. m. to7p. m 
21,497 
.955 
1 
546 
.591 
12.2 
272.4 
719.9 
4-6.6 
7 
p. m. to 1 a. m 
22,274 
1.055 
1 
273 
.218 
4.8 
282.1 
- .7 
1 
a. m. to 7 a. m 
22,173 
.752 
1 
090 
.338 
7.5 
247.3 
199.8 
-9.7 
Total 
91, 273 
29.7 
1,080. 9 
1,402.9 
- .1 
2, 513.4 
2 
7 
a. m. to 1 p. m 
23, 364 
.806 
1 
129 
.323 
7.5 
258.3 
445.5 
4-5.3 
1 
p. m. to 7 p. m 
23,356 
.938 
1 
269 
.331 
7.7 
284.2 
556.0 
4-3.0 
7 
p. m. to 1 a. m 
23,999 
.914 
1 
335 
.421 
10.1 
298.9 
- .3 
1 
a. m. to 7 a. m 
23, 541 
.777 
1 
089 
.312 
7.2 
267.9 
81.0 
-6.9 
Total 
94, 260 
32.5 
1,109.3 
1, 082.5 
4-1-1 
2,225.4 
3 
7 
a. m. tolp. m 
23,147 
.824 
1 
352 
.528 
12.1 
264.3 
504.6 
-6.0 
1 
p. m. to7 p.m 
23, 044 
.900 
1 
371 
.471 
10.8 
281.6 
575.5 
4-9.8 
7 
p. m. to 1 a.m 
22,423 
.710 
1 
164 
. 454 
10.2 
280.9 
- .7 
1 
a. in. to 7 a. m 
23, 344 
.848 
1 
171 
.323 
7,6 
273.1 
85.3 
-3.3 
Total 
91, 958 
40.7 
1, 099.9 
1,165. 4 
- .2 
2,305. 8 
4 
7 
a. m. tolp. m 
23, 268 
.798 
1 
134 
.336 
7.8 
258.7 
251.9 
4-5.8 
1 
p. m. to7p.m 
23, 031 
1.124 
1 
361 
.237 
5.5 
280.4 
560.6 
-6.7 
7 
p. m. to 1 a. m 
23, 833 
.793 
1 
123 
.330 
7.9 
295.0 
-3.6 
1 
a.m. to7 a.m 
25, 757 
.870 
1 
042 
.172 
4.4 
291.7 
118.0 
-3.1 
Total 
95, 889 
.... 
25.6 
1,125. 8 
930.5 
-7.6 
2,074.3 
Total, four days. 
373,380 
.... 
128.5 
4,415.9 
4,581.3 
9,118. 9 
Avg. per day ... 
93,345 
2, 279. 7 
Table 27.-Record of water in work experimen t (No. 6). 
The calorimetric results in the two experiments are shown in Tables 
28 and 29. The details of the table and methods of computation are 
as explained above in the description of the test experiments in which 
alcohol was burned in the respiration chamber. 
Table 28.-Summary of calorimetric measurements-Rest experiment (No. 9). 
Day. 
Period. 
Heat 
meas- 
ured. 
Change 
of 
tem- 
pera- 
ture of 
calorim- 
eter. 
Capac- 
ity 
correc- 
tion. 
Correc- 
tion due 
to tem- 
perature 
of food 
and 
dishes. 
Corrected 
heat. 
Water 
vapor- 
ized. 
Equiva- 
lent heat 
of water 
vapor- 
ized. 
Total 
heat. 
Cals. 
°C 
Cals. 
Cals. 
Cals. 
Grains. 
Cals. 
Cals. 
1 
7 a. m. to 1 p.m. 
517.9 
+0.12 
+ 7.2 
- 22. 2 
532. 9 
217.9 
129.0 
661.9 
1 p. m. to 7 p. m. 
556.6 
- .12 
- 7.2 
- 37.5 
511.9 
241.7 
143.1 
655.0 
7 p.m. to 1 a.m. 
472.7 
- .05 
- 3.0 
0 
469.7 
248.6 
147.1 
616.8 
1 a. m. to 7 a. m. 
277.3 
- .03 
- 1.8 
0 
275.5 
233.7 
138.4 
413.9 
Total 
1, 854.5 
-4.8 
- 59.7 
1, 790.0 
557.6 
2, 347. 6 
2 
7 a.m. tolp.m. 
490.7 
+ .04 
+ 2.4 
- 22.9 
470.2 
224.2 
132.7 
602.9 
1 p.m. to 7 p.m. 
489.2 
+ .08 
+ 4.8 
- 28.3 
465.7 
224.5 
132.9 
598.6 
7 p.m. to 1a.m. 
497.4 
- .15 
- 9.0 
0 
488.4 
243.3 
144.1 
632.5 
1 a. m. to 7 a. m. 
289.2 
+ .15 
+ 9.0 
0 
298.2 
221.3 
131.0 
429.2 
Total 
1, 766. 5 
- 
+ 7.2 
- 51.2 
1, 722. 5 
540.7 
2, 263. 2 
17951-No. 63 6 82 
Table 28.-Summary of calorimetric measurements-Best experiment (No. 9)-Cont'd. 
Day. 
Period. 
Heat 
meas- 
ured. 
Change 
of 
tem- 
pera- 
ture of 
calorim- 
eter. 
Capac- 
ity 
correc- 
tion. 
Correc- 
tion due 
to tem- 
perature 
of food 
and 
dishes. 
Corrected 
heat. 
Water 
vapor- 
ized. 
Equiva- 
lent heat 
of water 
vapor- 
ized. 
Total 
heal. 
Cals. 
°C 
Cals. 
Cals. 
Cals. 
Grams. 
Cals. 
Cals. 
3 
7 a. m. to 1 p. m. 
524.6 
+ .08 
+ 4.8 
- 21.2 
508.2 
216.7 
128.3 
636. 5 
1 p. m. to 7 p. m. 
512.1 
+ .08 
+ 4.8 
- 31.8 
485.1 
219.8 
130.1 
615.2 
7 p. m. to 1 a. m. 
486.6 
- .18 
-10.8 
0 
475.8 
239.7 
141.9 
617.7 
1 a. m. to 7 a. m. 
296.1 
+ .10 
+ 6.0 
0 
302.1 
220.9 
130.8 
432.9 
Total 
1,819.4 
+ 4.8 
- 53.0 
1,771.2 
531.1 
2, 302. 3 
4 
7 a. m. to 1 p. m. 
503.3 
- .05 
- 3.0 
- 21.3 
479.0 
222.4 
131.7 
610.7 
1 p.m. to 7 p. m. 
536.4 
+ .05 
+ 3.0 
- 29.2 
510.2 
220.3 
130.4 
640.6 
7 p. m. to 1 a. m. 
497.5 
- .10 
- 6.0 
0 
491.5 
244.7 
144.9 
636.4 
1 a. m. to 7 a. m. 
290.7 
+ .04 
+ 2.4 
0 
293.1 
245.0 
145.0 
438.1 
Total 
1,827. 9 
3.6 
50. 5 
1,773. 8 
552.0 
2, 325. 8 
Total, four 
days 
7, 268.3 
4-3.6 
214.4 
7, 057.5 
2,181. 4 
9, 238. 9 
Avg. per day 
1,817.1 
+ .9 
- 53 6 
1,764.4 
545.3 
2, 309. 7 
Table 29.-Summary of calorimetric measurements-Work experiment {No. 6). 
CO IO H* 
Day. 
Total 
Total, four days. 
Avg. per day ... 
S3 
5 
0 
S3 
3 
Total 
7 a. m. to 1 p. m 
1p.m. to7p. m 
7 n. m. to 1 am 
7 a. m.tolp. m 
1 p. m. to 7 p. m 
7 p. m. tol a. m 
1 a. m. to 7 a. m 
0 
S3 
JO 
s 
© 
B 
7 a. m. tol p.m 
Ip.m. to 7 p. m 
7n.m. am. . 
s' 
£ 
3 
» 
3 
7 a. m. to 1 p. m 
1 p.m. to 7 p.m 
7 n. m. tn 1 a,, m 
Period. 
2,897.6 
12, 357. 7 
3,089.4 
1,018.6 
1,059. 7 
521.7 
297.6 
3,091.1 
1,119. 6 
1,121.2 
588.8 
261.5 
© 
to 
JI 
to 
1,023.4 
1,153.9 
588.1 
259.8 
CO 
co 
& 
Cals. 
1,157. 3 
1, 270. 7 
617.7 
298.1 
Heat measured. 
1 
s 
+ + 
+ + 
* © 00 
1 
00 
1 + + 1 
O O 
to © © co 
1 + + 1 
-1 W CO 
1 
+ + 
o o >-* 
ji © u> 
1 
s 
°C 
+0.50 
- .10 
- .25 
Change of tem- 
perature of cal- 
orimeter. 
; + + 
-0 
£ 
© 
+ + 
© co © 
+ 
© 
© 
1 + $ 
-* CO g* 
Ji © © • 
Capacity correc- 
tion. 
- 29.4 
149. 5 
| 10.2 
- 47. 9 3, 035. 2 
- 16.0 1, 095. 6 
- 31.9 1,092.3 
592.8 
254.5 
X 
1 1 
to 
to © 
© 
JI 
Cals. 
- 11.7 
- 28.8 
Correction due to 
temperature of 
food and dishes. 
co 
o 
Ui 
co 
00 
2,875. 2 
12 215. 2 
1,016. 4 
1, 046.5 
517.7 
294.6 
to 
© 
© 
to 
Ji 
1,024.9 
1,132. 7 
585.1 
249.8 
3,312.31, 110.5 
Cals. | Gms. 
1,175.6 284. 3 
1, 235. 9 284.5 
602. 7 286.9 
298.1' 254.8 
Corrected heat. 
1,151.4 681. 6 
4, 544.32,690.1 
1,136.1 672.5 
266.5 157. 8 
285.9' 169.2 
302.9 179.3 
296.1 175.3 
1,140. 6 675. 2 
276.4 163.6 
292. 5 173.2 
291.1 172.3 
280. 6 166.1 
X 
265.8 
291.9 
309.0 
275.1 
Water vaporized. 
© 
JI 
157.3 1,182.2 
172.8 1,305.5 
182.9 768. 0 
162. 9 412.7 
657.4 3,969.7 
Cals. 
168.3 
168.4 
169.9 
150.8 
Equivalent / heat 
of water vapo- 
rized. 
3 
© 
3,556.8 
14 905. 3 
1,174. 2 
1,215.7 
697.0 
469.9 
3,710.4 
1, 259.2 
1, 265.5 
765.1 
420.6 
i 
00 
© 
Cals. 
1, 343. 9 
1,404.3 
772. 6 
Total heat. 
224 
240 
WK 
9EG 
210 
210 
Mins. 
238 
216 
Dura- 
| tion. 
Work done. 
1 255 
1,023 
256 
39 125 
38 130 
10 
£ 
40 135 
38 133 
39 
39 
Watts 
42 
41 
Rate. 
o 
o 
115 
115 
to 
© 
Cals. 
143 
127 
Heat equivalent. 83 
As has been previously mentioned, the urine was collected in six- 
hour periods beginning at 7 a. m. (the hour of beginning and ending 
the experiment). The amount of nitrogen in the urine from 7 a. m. of 
one day to 7 a. m. next day is taken as a measure of the protein metab- 
olized during this period. Of course this makes no allowance for the 
nitrogen lag, but, as explained on page 93, it was considered that the 
error thus introduced might be ignored, since the subject had been living 
on the same diet and had had the same exercise from day to day 
for the experimental period of four days and the previous four days of 
the preliminary experiment also. The amount of nitrogen in the urine 
for 12 to 24 hours after the close of the experiment was also determined 
in six-hour periods, although these values are not recorded here. In 
Table 30 is summarized the amount of urine for each day of the rest 
and the work experiments, together with the percentage and amounts 
of nitrogen and carbon and the heat of combustion per gram, and the 
total heat of combustion of the urine. 
Table 30.-Amount and composition of urine. 
Experiment and day. 
Amount. 
Nitrogen. 
Carbon. 
Heat of combustion. 
Per gram. 
Total. 
Rest experiment (No. 9): 
Grams. 
Per cent. 
Grams. 
Per cent. 
Grams. 
Calories. 
Calories. 
First day 
1, 855. 3 
1.01 
18. 75 
0.69 
12.80 
0.082 
152 
Second day 
1, 977. 6 
.95 
18. 75 
.65 
12.79 
.081 
160 
Third day 
1,510.6 
1.21 
18.28 
.83 
12. 50 
.095 
143 
Fourth day 
1, 358. 9 
1.32 
17.89 
.90 
12.19 
.102 
139 
Total, four days 
6, 702.4 
73.68 
50. 28 
594 
Average per day 
1, 675.6 
18.42 
12. 57 
149 
Work experiment (No. 6): 
First day 
1,349.9 
1.27 
17.17 
.97 
13.12 
.094 
127 
Second day 
1,109.4 
1.48 
16. 39 
1.13 
12.54 
.119 
132 
Third day 
1,033.8 
1.47 
15.22 
1.13 
11. 72 
.115 
119 
Fourth day 
1,304. 7 
1.25 
16. 34 
.96 
12. 50 
.094 
123 
Total, four days 
4, 797. 8 
65.12 
49. 88 
501 
Average per day 
1,199. 4 
16.28 
12.47 
125 
The data previously given serve for the calculation of Tables 31 and 
32, which show the gain or loss of nitrogen, carbon, protein, and fat 
and the comparison of the estimated heat of the material oxidized in 
the body with the heat actually measured in the rest and work experi- 
ments. The calculations are partially explained by the letters and 
algebraic formula at the top of the columns. Thus column d indicates 
the gain or loss of nitrogen, and is computed by adding the amounts in 
the feces and urine, as shown in columns b and c, and subtracting this 
sum from the amount in the food as shown in column a.1 In Table 32 
1 For further explanations of these calculations see U. S. Dept. Agr., Office of Exper- 
iment Stations Bui. 44. 84 
the values given in columns n, o, and p are actual heats of combus- 
tion as determined by the bomb calorimeter. The values in columns q 
and r are obtained by the use of factors representing the average heat 
of combustion of I gram of protein and fat, respectively. The values 
in column t are obtained from Tables 28 and 29. 
Table 31.-Income and outgo of nitrogen and carbon, with estimated gain or loss of 
protein and fat. 
Experiment 
and day. 
Nitrogen. 
e. 
Pro- 
tein 
gain 
(+) 
loss 
(-) 
- d x 
6.25. 
Carbon. 
k. 
Car- 
bon in 
pro- 
tein 
gained 
(+) 
or lost 
(-) 
-e X 
0.53. 
I. 
Car- 
bon in 
fat 
gained 
( + ) 
or lost 
(-) 
=j-k. 
m. 
Fat 
gain 
(+) 
loss 
(-) 
=1 + 
0. 765. 
a. 
In 
food. 
b. 
In 
feces. 
c. 
In 
urine.1 
d. 
Gain 
( + ) 
loss 
(-) 
- a - 
(6+c). 
f. 
in 
food. 
fc 
feces. 
h. 
In 
urine. 
i. 
In 
respir- 
atory 
prod- 
ucts. 
j- 
Gain 
( + ) 
loss 
(-) 
=f- 
(g + h 
+ 0- 
Rest experi- 
ment (No. 9). 
Gms. 
Gms. 
Gms. 
Gms. 
Gms. 
Gms. 
Gms. 
Gms. 
Gms. 
Gms. 
Gms. 
Gms. 
Gms. 
First day .... 
19.1 
1.2 
18.7 
-0.8 
- 5.0 
261.5 
13.3 
12.8 
226.7 
+ 8.7 
- 2.7 
+ 11.4 
+ 14.9 
Second day... 
19.1 
1.3 
18.8 
-1.0 
- 6.3 
261.6 
13.4 
12.8 
221.8 
+ 13.6 
- 3.3 
+ 16.9 + 22.1 
Third day.... 
19.1 
1.2 
18.3 
- .4 
-2.5 
261.5 
13.3 
12.5 
220.7 
+ 15.0 
- 1.3 
+ 16.3 + 21.3 
Fourth day... 
19.1 
1.3 
17.9 
- .1 
- 0.6 
261.6 
13.4 
12.2 
225.3 
+ 10.7 
- .3 
+ 11.0j+ 14.4 
Total 
76.4 
5.0 
73.7 
-2.3 
-14.4,1,046. 2 
53.4 
50.3 
894.5 
+ 48.0 
- 7.6 
+ 55.6 + 72.7 
Average per 
day 
19.1 
1.2 
18.4 
- .6 
- 3.6 
261.5 
13.3 
12.6 
223.6 
+ 12.0 
- 1.9 
+ 13.9 
+ 18.2 
Work experi- 
ment (No. 6). 
First day 
19.1 
1.5 
17.5 
+0.1 
+ 0.6 
336.7 
12.3 
13.1 
364.0 
- 52.7 
+ 0.3 
- 53.0 
- 69.3 
Second day... 
19.1 
1.5 
16.6 
+ 1-0 
+ 6.2 
336.7 
12.4 
12.6 
342.1 
- 30.4 
+ 3.3 
- 33.7 
- 44.1 
Third day.... 
19.1 
1.5 
15.4 
+2.2 
+ 13.8 
336.7 
12.3 
11.7 
345. 2 
- 32.5 + 7.3 
- 39.8 
- 52.0 
Fourth day... 
19.1 
1.5 
16.5 
+ 1-1 
+ 6.9 
336.7 
12.4 
12.5 
329.4 
- 17.6 + 3. 7 
- 21. 3 
- 27.8 
Total 
76.4 
6.0 
66.0 
+4.4 
+27. 5 1,346.8 
49.4 
49.91,380.7 
-133.2 
+ 14.6 
-147. 8 
-193. 2 
Average per 
day 
19.1 
1.5 
16.5 
+ 1.1 
+ 6.9 
336.7 
12.4 
12.5 
345.2 
- 33. 3 + 3. 6 
- 36.9 
- 48.3 
'Including nitrogen of perspiration. 85 
Table 32.-Income and outgo of energy. 
Experiment and day. 
Energy (heats of combustion). 
t. 
Heat 
actual- 
ly meas 
ured. 
Ratio of 
beat 
measured 
to that of 
material 
actually 
oxidized 
t-r-s. 
Actually determined. 
Estimated. 
n. 
Of food. 
0. 
Of feces. 
P- 
Of urine. 
1- 
Of pro- 
tein gain- 
ed ( + lor 
lost (-) 
- 3X5.5. 
Of fat 
gained(-f-) 
or lost(-) 
=mX94. 
s. 
Of mate 
rial actu- 
ally oxi- 
dized in 
body 
- n -(o+ 
p+q+r). 
„ ..... 
Rest experiment 
(No. 9). 
Calories. 
2,717 
2, 717 
2,717 
2,717 
Calories. 
142 
142 
142 
142 
Calories. 
152 
160 
143 
139 
Calories. 
- 27 
- 35 
- 14 
- 3 
Calories. 
+ 140 
+ 208 
+ 200 
+ 135 
Calories. 
2, 310 
2, 242 
2,246 
2, 304 
Calories 
2,348 
2, 263 
2,302 
2,326 
Per cent. 
Total 
10,868 
2,717 
568 
142 
594 
149 
- 79 
- 20 
+ 683 
+ 171 
9,102 
2, 275 
9, 239 
2, 310 
Avg. per day . 
Work experiment 
(No. 6). 
First day 
101.5 
3, 678 
3,678 
3, 678 
3,678 
139 
139 
139 
139 
127 
132 
119 
123 
+ 3 
+ 34 
+ 76 
+ 38 
- 651 
- 415 
- 489 
- 261 
4,060 
3, 788 
3,833 
3, 639 
3, 970 
3.668 
3.710 
3, 557 
Second day . _ 
Third day 
Fourth day 
Total 
Avg. per day... 
14,712 
3,678 
556 
139 
501 
125 
+151 
+ 38 
-1,816 
- 454 
15, 320 
3, 830 
14, 905 
3, 726 
97.3 
SUMMARY OF RESULTS OF EXPERIMENTS WITH A MAN. 
The results of the two experiments reported above are summarized 
in the following tables. Table 33 shows the balance of income and 
outgo of nitrogen and carbon, and the calculated gain or loss of protein 
and fat. Table 34 shows the daily income and outgo of protein and 
energy. 
Table 33.-Average daily income and outgo of nitrogen and carbon in the rest and work 
experiments (Nos. 9 and 6), with the estimated gain or loss of protein and of fat. 
Experiment. 
Nitrogen. 
Carbon. 
Calculated 
gain or loss. 
In 
food. 
In 
feces. 
In 
urine. 
Gain 
(+) or 
loss 
(-)• 
In food. 
In 
feces. 
In 
urine. 
In re- 
spira- 
tory 
prod- 
ucts. 
Gain 
(+) or 
loss 
(-)■ 
Of 
pro- 
tein. 
Of 
fat. 
Gms. 
Gms. 
Gms. 
Gms. 
Gms. 
Gms. 
Gms. 
Gms. 
Gms. 
Gms. 
Gms. 
Rest (No. 9) 
19.1 
1.2 
18.4 
- 0.6 
261.5 
13.3 
12.6 
223.6 
+ 12.0 
-3.6 
+ 18.2 
Work (No. 6) 
19.1 
1.5 
16.5 
+ 1.1 
336.7 
12.4 
12. 5 
345.2 
-33.3 
+6.9 
-48.3 86 
Table 34.-Comparison of daily income and outgo of protein and of energy in the rest 
and work experiments {Nos. 9 and 6). 
Experiment. 
Protein. 
Energy. 
Of 
food. 
Actual- 
ly oxi- 
dized. 
Of 
food. 
Of ma- 
terial 
actual- 
ly oxi- 
dized. 
Meas- 
ured. 
Differ- 
ence in 
percent 
of heat 
of ma- 
terial 
actual- 
ly oxi- 
dized. 
Gms. 
Gms. 
Cals. 
Cals. 
Cals. 
Per cent. 
Rest (No. 9) 
119.4 
115.0 
2, 275 
2,354 
2,310 
4-1.5 
Work (No. 6) 
119.4 
103.1 
3,830 
3,864 
3, 726 
-2.7 
The difference between income and outgo of energy in experiment 
No. 9 amounts to 1.5 per cent and in experiment No. 6 to 2.7 per cent 
of the energy of the material oxidized. This discrepancy may be due 
to experimental errors, of which several are possible, as pointed out on 
pages 90 to 94. 
There was one source of error of special consequence in experiment 
No. 6. The muscular work of the subject was at times rather severe. 
As a consequence, heat was developed within the apparatus at a rapid 
rate and the changes in temperature within the chamber were con- 
siderable. We are inclined to think that the heat measurements under 
these circumstances were less accurate than usual, and that minor 
modifications of the apparatus may be needed to provide for greater 
accuracy in experiments of this class. 
In later experiments, which are not yet reported, the differences 
between income and outgo of energy range much nearer to the theo- 
retical; in other words, in these more accurate experiments not far 
from 99 per cent of the potential energy of the material metabolized 
and oxidized in the body is accounted for in the kinetic energy given 
off in the forms of heat and external muscular work. 
In how far this fairly close agreement is due to a counterbalancing 
of errors it is impossible to say. But in view of the physiological diffi- 
culties in the way of absolutely accurate results, and the evident 
possibility of minor errors in the purely chemical and physical deter- 
minations, and likewise in the factors used for computation, this agree- 
ment seems to us very satisfactory. We believe that it marks a stage 
in the development of the apparatus and methods sufficient to war- 
rant extended series of experiments upon various questions connected 
with the laws of nutrition, and such experiments have been begun. 
Efforts are being made at the same time to eliminate part at least of 
the experimental errors. At present these efforts are chiefly in the 
direction of improvement of methods of sampling and analyzing the 
food materials and excretory products, and finding of minor sources 
of error in the determination of carbon, hydrogen, and heat given off 
in the respiration chamber, and the direct determination of the oxy- CORRECTION. 
Please correct figures in the third and fourth columns of Table 34, 
p. 86, Bulletin 63, Office of Experiment Stations, United States 
Department of Agriculture, as follows: 
Of 
food. 
Of ma- 
terial 
actual- 
ly oxi- 
dized. 
Cals. 
Cals. 
2,717 
2,275 
3,678 
3,830 87 
gen of income and outgo. Minor alterations are also being made in 
the apparatus and the methods of its manipulation by which it is 
hoped that somewhat greater accuracy may be secured. 
GENERAL SUMMARY. 
The attempt has been made in the preceding pages to describe-(1) a 
new form of respiration calorimeter and the methods of its use, and (2) 
several experiments in which the apparatus and methods have been 
employed. 
The experiments here described had a twofold purpose: To test the 
accuracy of the apparatus and methods, and to determine the balance 
of income and outgo of matter and energy in the body. They are pre- 
liminary to more extended research upon some of the fundamental 
problems of nutrition. 
The name here used for the apparatus, "respiration calorimeter," is 
suggested by the fact that it is essentially a respiration apparatus with 
appliances for calorimetric measurements. As a respiration apparatus 
it is similar in principle to that of Pettenkofer. The calorimeter is 
essentially a water calorimeter; that is to say, the heat evolved in the 
chamber is measured by a current of cold water. The appliances for 
measurement of both the respiratory products and the heat given off 
from the body differ in important respects from those of any other 
apparatus with which we are familiar. 
The most important feature of the respiration calorimeter is a metal- 
walled chamber in which a man lives, eats, drinks, works, and sleeps. 
Provision is made for ventilating the chamber and for regulating the 
temperature and moisture of the air within it. The volume of the 
ventilating air current is measured and samples for analysis are taken 
before and after it passes through the chamber, thus obtaining the 
amounts of carbon dioxid and water in the respiratory products. The 
food, drink, feces, and urine are weighed and analyzed, and their 
potential energy is determined, as is the kinetic energy given off from 
the body in the forms of heat and external muscular work. 
The devices for measuring the heat given off from the body include: 
(1) Arrangements to prevent gain or loss of heat in the chamber either 
by the passage of heat through the walls or the bringing in and taking 
out of heat in the ventilating air current; (2) arrangements by which 
the heat given off in the chamber, by the body or otherwise, is carried 
out by the current of water above referred to. This current, which is 
conveyed by copper pipes, comes into the chamber at a low tem- 
perature, passes around the interior, absorbs the heat, and goes out 
correspondingly warmer. The quantity of the water and the rise of 
temperature show how much heat is carried out. The measurements 
of temperature of the inner walls of the apparatus, the air inside, and 
the incoming and outgoing air and water currents are made chiefly by 
electrical means and are very delicate, differences of one-hundredth of 88 
a degree being easily determined. By regulating the temperature of 
the water current when it enters the chamber and its rate of flow, the 
temperature of the interior air and walls can be kept almost exactly 
constant. When slight changes of interior temperature do occur the 
corresponding amounts of heat taken up or given off by the interior 
air and parts of the apparatus are readily determined. Applying the 
corrections, generally very small, for such changes of interior tempera- 
ture, the heat gained by the water current in passing through the 
chamber is the measure of the amount of heat developed in the 
chamber. 
The accuracy of the apparatus and of the methods for the determi- 
nations of carbon dioxid, water, and heat was tested by heat generated 
in the chamber bypassing an electric current through a resistance coil 
and also by burning ethyl alcohol within the chamber. In the electrical 
tests the measurements of heat generated and found were practically 
identical. In the alcohol tests the average amounts found by actual 
experiment were: For carbon, 99.9 per cent; hydrogen, 100.6 per cent; 
and heat, 99.9 per cent of the theoretical amounts. It thus appears 
that this apparatus when used for the analysis of alcohol and the deter- 
mination of its heat of combustion gives results nearly, if not quite, as 
accurate as are obtained by the ordinary laboratory methods which can 
be used only with small amounts. 
The measurements of heat given off from the body of a man inside 
the chamber are so delicate that very slight bodily movements, such as 
rising from a chair or turning over in bed, are immediately noticed 
by the observer, who is constantly watching the galvanometer and 
thermometers. 
The experiments with men, two of the earlier of which are here 
reported, were undertaken for the study of several problems. The 
question especially considered here is this: Is the energy given off from 
the body in the form of heat, or of heat and external muscular work, 
equal to the potential energy or heat of combustion of the material 
actually burned in the body? In other words, when the compounds of 
the food and the body-proteids, fats, and carbohydrates-are burned 
is their potential energy transformed into the equivalent kinetic energy 
and-into forms which can be measured by the means here used? Or, 
to state the question more broadly, does the law of the conservation 
of energy obtain in the living organism? 
The experiments with a man generally continued eight days, during 
the last four days and five nights of which the subject was in the 
respiration chamber. The diet during each experiment was uniform 
throughout the whole eight days. The purpose of the preliminary period 
of four days was to bring the body into at least approximate nitrogen 
and carbon equilibrium with the food and to make the determination 
of the amounts of nutrients absorbed as nearly accurate as practicable. 
The income and outgo of nitrogen were determined during this period, 89 
which thus amounted to a digestion and metabolism experiment. The 
metabolism of nitrogen, carbon, hydrogen, and energy was determined 
during the final period of four days. 
Two such experiments are here described. In the first of these, 
called a "rest experiment," the man had as little muscular exercise as 
he could well have with comfort. In the second, called a "work experi- 
ment," he was engaged in quite active muscular exercise. The exter- 
nal muscular work was expended in driving a dynamo which produced 
an electric current. The latter was measured, thus showing the amount 
of external muscular work performed, and was passed through a resist- 
ance coil and the energy was transformed into heat, which was meas- 
ured with that given off from the body. 
The difference between the income and outgo of energy as measured 
in these two cases was +1.5 and -2.7 per cent, and averaged -0.6 per 
cent of the income. The amount of energy as measured was in one case 
more and in one case less than the theoretical amount of potential 
energy in the material consumed. The larger discrepancy was in 
experiment No. 6. Certain sources of error in this appear to have been 
eliminated, at least in part, in later experiments, of which experiment 
No. 9 was one. In these later experiments the agreement is very close, 
the energy measured being about 99 per cent of the theoretical. On 
the whole, the agreement between theoretical amounts of energy trans- 
formed and those found in the experiments is as close as could be 
reasonably expected, in view of the physiological sources of experi- 
mental error and the uncertainties as to the correctness of some of the 
physical and chemical constants used as factors in the calculations of 
results. 
A gradual approach toward a demonstration of the application of the 
law of the conservation of energy in the living organism has been 
made in the experimental inquiry of later years. This is notably the 
case with Rubner's1 numerous experiments with two dogs in a specially 
devised respiration calorimeter. In some of the experiments the dogs 
fasted; in others they had lean meat or fat bacon, or both, for food. 
The actual determinations were: The weight of the animal at the 
beginning and end of the experiment and the weights of food and 
water given and of feces and urine excreted; the percentages of fat in 
the bacon fed and of nitrogen in the feces and urine; the weights of 
carbon dioxid and water in the respiratory products, and the amounts 
of heat given off from the body. No balance of nitrogen or carbon was 
determined, but equilibrium of each was assumed. The amounts of 
food and body materials oxidized were calculated from the nitrogen 
and carbon excreted. The energy of income was calculated from the 
estimated amounts of compounds oxidized in the body, and the outgo 
was taken as the heat given off from the body. Nine experiments 
made in this way gave values ranging from about -5.2 to -|-3.2 per cent 
'Die Quelle der thierischen Warme, Ztschr. Biol., 30 (1894), pp. 73-86. 90 
of the total energy as computed, and the average for the whole forty- 
five days of the experiments showed a difference of less than one-half 
of 1 per cent between the estimated potential energy of food materials 
calculated to have been oxidized and the heat given off from the body 
as measured by the calorimeter. 
In the opinion of the author, these experiments prove that the nutri- 
ents of the food and the body materials consumed are the sole sources 
of heat in the animal body. Although the actually determined data 
of the experiments are so few and the estimates and assumptions so 
many, there can hardly be a doubt that the opinion is justifiable. The 
experiments thus tend to confirm the belief that the law of conserva- 
tion of energy applies in the living organism. The same may be said 
of the experiments of Studenski, referred to on page 11, even though 
the data of his experiments, like those of Rubner, have not all the com- 
pleteness that is to be desired. 
In the experiments described in this bulletin, the energy of outgo 
includes both the heat given off from the body and, in one case, that of 
a considerable amount of external muscular work. 
In the experiment No. 9 (rest), the outgo apparently exceeds the 
income by about 35 calories per day. This quantity is really quite small. 
It would correspond to the potential energy of about 1 grams of body 
fat, or nearly the same weight of butter, or 9 grams of sugar, or 14 
grams of bread. 
It may be that a small part of the energy which is transformed in 
the body is given off in some form which the apparatus and methods 
here used are incapable of measuring. The income and outgo of energy 
agree as closely, perhaps, as could be expected in view of the possible 
sources of error and defects in apparatus and methods. 
The sources of error and uncertainty in these experiments can, it 
seems to us, be divided into two classes; those incident to the experi- 
mental work as such, and those due to uncertainty as to the physical, 
chemical, and physiological constants used as factors in the computa- 
tions. 
The chief sources of experimental uncertainty and error appear to be 
the following: 
(1) Appliances for the direct determination of oxygen of income and 
outgo are still lacking. These determinations of oxygen are needed 
for aid in estimating the amounts of carbohydrates, water, and oxygen 
gained or lost by the body during the experiment. Efforts to supply 
this deficiency are being made. 
(2) No determinations were made of the amounts of sulphur, phos- 
phorus, and chlorin of income and outgo. These more complete chemi- 
cal determinations are needed to throw light upon the character of the 
proteids and other nitrogenous compounds burned and stored or lost 
in the body. 
(3) No provision has yet been made for the measurement and analysis 91 
of the volatile organic compounds given off from the body, for example, 
the gaseous products of bacterial action in the intestines. Research 
elsewhere, however, has shown them to be extremely small,1 and it is 
hardly probable that they would materially affect the results. 
(4) The ordinary methods for the sampling and analysis of food and 
excretory products are not as accurate as is to be desired. Considerable 
attention has been given to this subject in connection with the present 
investigation. 
(5) Among the physical constants used in computing the results of 
the experiments the following are of more or less uncertain accuracy. 
The factor 0.592 is here employed for the latent heat of vaporiza- 
tion of water at 20° C, which is about the usual temperature of the 
chamber. This factor is derived from the formula of Regnault and is 
apparently not very far from the truth. The exact value, however, 
remains to be determined. Furthermore, the conditions of vaporization 
in the respiration chamber are different from those of the experimental 
determinations upon which the value 0.592 is based. It is not impossi- 
ble that accurate determinations of the latent heat of vaporization of 
water under the conditions that obtain in the chamber of the respira- 
tion calorimeter may call for a change in this factor sufficient to meas- 
ureably affect the results. 
(6) The heat of combustion of ethyl alcohol is here taken as 7.067 
calories per gram, the average of a considerable number of determina- 
tions made in this laboratory. It agrees quite closely with Berthelot's 
figure 7.079, which is the average of but two determinations. Berthe- 
lot's determinations were made with a bomb calorimeter of his own 
invention, ours with a modified form used in this laboratory. Exact 
determinations of the heat of combustion of such substances are diffi- 
cult and the results of individual determinations differ considerably. 
Efforts are now being made to learn somewhat more of the sources of 
error, in the hope of obtaining more certain results. It is of course 
possible that later experiments may call for a different factor. Mean- 
while it may be said that agreement of the results obtained by the 
bomb and respiration calorimeters tend to confirm the accuracy of 
both. 
The measurements of the income and outgo of heat in the respira- 
tion calorimeter by the electrical methods above described were practi- 
cally identical. This implies that the apparatus as a calorimeter gives 
very accurate results. It also tends to confirm the correctness of the 
determinations of the heat of combustion of alcohol. The inference 
would be that the discrepancies between the theoretical amounts of 
heat and those measured in the experiments with men must be due in 
1 See Billings, Mitchell, and Bergey, On the composition of expired air and its 
effects upon animal life: Washington, Smithsonian Institution, 1895; and Bergey, 
Methods for the determination of organic matter in air: Washington, Smithsonian 
Institution, 1896. 92 
part to other sources of error than those of the actual measurements 
by the calorimeter. 
Considering the respiration calorimeter as an apparatus for the 
analysis of alcohol and the determination of its heat of combustion on 
a large scale, the results obtained compare with the theoretical values 
for carbon and hydrogen and those obtained with the bomb calorimeter 
for heat of combustion, as in Table 35, which follows. The average 
figures in this table represent the totals of Table 19, p. 72. 
Table 35.-Analysis and determination of heat of combustion of alcohol by the respiration 
calorimeter. 
Alcohol test experiment. 
Carbon. 
Hydro- 
gen. 
Heat of 
combus- 
tion per 
gram. 
Per cent. 
1 er cent. 
Calories. 
No. 1 
52.17 
13. 08 
7.007 
No 2 
52.12 
13.04 
7.126 
No. 3 
52.85 
a [13. 84] 
7.013 
No 4 
a[50. 34] 
a |14.19] 
7.096 
No. 5 
52. 59 
13.16 
7. 069 
51.44 
13.21 
7. 030 
No. 7 
52.17 
13.20 
a [7.219] 
No. 8 
51. 91 
13.18 
7. 049 
No. 9 
51.86 
13.01 
7. 071 
Average 
52.12 
13.12 
7.062 
Theoretical; heat of combustion by bomb calorimeter 
52.17 
13.04 
7. 067 
a Omitted from averages on account of obvious errors. 
(7) There are numerous possibilities of slight errors in the physi- 
ological constants and assumptions used in the computations of the 
results obtained with a man in the respiration chamber of the calorim- 
eter. Of these the more important are the following: One of the 
chief sources of uncertainty in metabolism experiments with man and 
animals, in which the balance of income and outgo is determined, has 
to do with the quantity and composition of the materials actually 
absorbed from the food in the alimentary canal during the experi- 
mental period. The best that can be done here is to endeavor to have 
the diet and other conditions of the experiment as nearly uniform as 
possible for the time covering not only the experimental period but also 
several preceding days, and to make the experimental period as long 
as practicable. In the experiments here described the whole period is 
made eight days, of which the last four are given to the metabolism 
experiment proper. 
(8) The amounts of protein and fat gained or lost by the body during 
the period of actual experiment, namely, the four days which the sub- 
ject spends in the chamber, are computed from the observed gain or 
loss of nitrogen and carbon. In these computations it is assumed that 
the body gains or loses no compounds containing nitrogen and carbon 93 
except protein and fats. Any possible difference between the store of 
carbohydrates at the beginning and that at the end of the experiment 
is thus left out of account. In making the computations it is assumed 
that the protein compounds-that is proteids and other nitrogenous 
compounds-contain collectively 16 per cent of nitrogen and 53 per 
cent of carbon, and that the fats contain 76.5 per cent of carbon. These 
figures are arbitrary, and though based upon the best available data 
may differ considerably from the actual facts. The heats of combustion 
of the proteids and fats thus estimated to be gained or lost are taken 
as 5.5 calories and 9.4 calories per gram, respectively.1 These values, 
like the chemical constants just cited, are probably more or less incorrect, 
and it is not impossible that the errors introduced by their use may be 
sufficient to account for a considerable share of the discrepancy between 
the estimated income and the measured outgo in the later experiments 
with man. 
(9) Still another possibility of error is found in the so-called nitrogen 
lag in the urine; that is to say, in the period between the end of the 
experiment and the final excretion of the nitrogen which belongs to 
the experimental period. We have assumed that when the diet and 
muscular work have been performed for a period of eight days, of which 
the last four were spent in the respiration chamber, the amount of 
nitrogen in the urine collected from 7 a. m. to 1 a. m. is equivalent to 
the amount metabolized in the same period. While this assumption 
would probably be more or less incorrect for any given day, we have 
believed it would be not far out of the way on the average. We have, 
however, collected and analyzed the urine for each six hours during 
the experiment and for similar periods during the twelve or twenty- 
four hours after the end of the experiment. The results of these deter- 
minations have not yet been published. 
(10) Finally, a possible source of error is in changes of temperature 
of the body. If the body is warmer or colder at the end of the experi- 
mental period than at the beginning allowance should be made for the 
gain or loss of heat. The amount of this change of heat is dependent 
upon the change of temperature, and the specific heat of the body. 
To measure the temperature of the body exactly is not easy. The 
specific heat of the body is not precisely known nor could it at best be 
stated with precision in any given case, because it would depend upon 
the weight and specific heats of different constituents, as water, pro- 
teids, fats, and other compounds. This subject is receiving considera- 
tion in connection with the experiments. It will suffice to say here 
that measurements of body temperature now made with a thermometer 
by the ordinary clinical methods did not reveal any marked differences 
between the temperature at the hour when each experimental period 
began and that when it ended, namely, at 7 a. m. in each case. 
1 U. S. Dept. Agr., Office of Experiment Stations Bui. 21, pp. 126-133. 94 
In brief, so far as the sources of error are those of chemical and phys- 
ical analyses and manipulation, the attempt is being made to reduce 
them to a minimum. In so far as they are of physiological origin and 
practically beyond control the attempt is made to eliminate them so far 
as practicable by long experimental periods and by duplication of the 
experiments. 
Meanwhile it is safe to say that in view of the physiological uncer- 
tainties and sources of error and the probable incorrectness of some of 
the physical and chemical constants employed, the differences between 
the estimated income and measured outgo of energy in the experiments 
with men are not at all surprising. 
In view of these defects and sources of error in methods and appa- 
ratus we would perhaps be unwarranted in assuming that the experi- 
ments thus far made completely demonstrate the application of the law 
of the conservation of energy in the human organism. They do, how- 
ever, seem to us to be reasonably near to such demonstration, especially 
when we take into account the possible sources of error-chemical, 
physical, and physiological-which have been discussed above. 
It is certainly safe to assume that the principle followed in the 
experiments is correct, and that the apparatus and methods are accu- 
rate to the degree required for the experimental study of a large vari- 
ety of the fundamental problems of biological chemistry and physics. 
Among these are the metabolism of energy and the production of heat 
by the body in the performance of its ordinary functions, as circula- 
tion, respiration, and digestion; the relations of muscular and mental 
work to the metabolism of matter and energy; the demands of the 
body for nutriment under different conditions of work and rest; the 
duties performed by the different nutrients of food in supplying 
the needs of the body; and finally, the nutritive values of food mate- 
rials and the amount and proportions best adapted to the needs of 
people of different classes, with different occupations, and in different 
conditions of life. That such inquiries may be valuable for the study 
of food and nutrition in disease is equally apparent. Of course they 
are fundamentally necessary for a more thorough understanding of the 
economy of feeding domestic animals.