tinued after the work-periods ceased. Under these conditions the carbondioxide production and oxygen consumption were both measured, and careful records of the pulse-rate made. In a few experiments a record of the respiration was likewise made by means of the spirometer attached to the ordinary form of respiration apparatus, so as to obtain data regarding the respiratory types before and after work, but for reasons given subsequently it was impracticable to use the spirometer during the muscular-work periods.

DETERMINATION OF THE NITROGEN EXCRETION.

It was hoped that a reasonably careful study of the urine during severe muscular work could be made, although evidence in the earlier literature with regard to the nitrogen excretion is much more abundant than that with regard to the gaseous exchange. Unfortunately, while the subject was most tractable and coöperated in every way with the investigation while inside the laboratory, it was very difficult to control the diet outside and even more difficult to insure a proper collection of the urine and feces; hence we soon had to give up the plan of securing 24-hour specimens of urine for comparison purposes. When possible, however, the subject emptied the bladder before the experiment began and again after it was over. The data secured have been collected and will be subsequently discussed, but they give only fragmentary evidence with regard to the nitrogen excretion as affected by severe muscular work.

PULSE OBSERVATIONS.

The intimate relationship between the pulse-rate and the metabolism so regularly noticed in this laboratory led to an attempt to secure the pulserate during these experiments. When the subject was lying quietly upon the couch, the observations were made with the greatest fidelity and regularity and became an integral part of each individual period. When the subject was riding on the ergometer, it was very difficult to secure accurate records of the pulse-rate, although attempts were made to obtain these by means of the stethoscope and occasionally from the radial pulse. While observations were made in practically every experiment, they were by no means so frequent as we should have liked, and obviously some other form of technique will be essential for studying satisfactorily the pulse-rate under these conditions. At the time the experiments were made, we had no string galvanometer or oscillograph at our disposal for securing such records.

DETERMINATION OF THE ALVEOLAR AIR.

The marked changes in the respiratory quotient noted in many experiments led to observations on the alveolar air, and the influence of muscular work upon its composition. These determinations were made through the assistance of Mr. H. L. Higgins of the laboratory staff, and were carried out by the Haldane method, the sample being analyzed by means of the small Haldane gas-analysis apparatus."

a

OBSERVATIONS ON THE INFLUENCE OF DIET.

The difficulties incidental to an adequate control of the subject while outside the laboratory and his disinclination to follow rigidly a diet, even

a Haldane and Priestley, Journ. Physiol., 1905, 32,p. 225.

Haldane, Methods of air analysis, London, 1912.

though this was made as liberal as possible, rendered it impracticable to secure much information regarding the effect of diet upon the efficiency of the body as a machine or upon the character of the katabolism. On the other hand in a series of experiments which was carried out with reasonable care, a certain amount of evidence was obtained which can be used with some reserve for discussing this point. Naturally the ideal arrangement would have been to make all of the experiments during a period in which the diet could be carefully controlled for several days previous, but as this was impossible, the plan for a detailed study of the influence of diet upon the character of the metabolism had to be abandoned in order to retain the services of an exceptionally good subject who coöperated in other ways.

EXPERIMENTS WITH UNTRAINED SUBJECTS.

The majority of the experiments reported were made with a trained. athlete, and were conducted over such a long period that differences, if any, which might be due to the influence of season, weather, diet, the alterations in training, the effect of over-training, and psychological condition could be noted. It was considered desirable, however, to make additional observations on a number of untrained individuals. These men, most of whom were members of the laboratory staff, kindly volunteered to subject themselves to these severe tests, even at the risk of becoming very muscle-sore. The evidence secured with the untrained subjects is especially valuable when considered in comparison with the data obtained in the experiments with the trained athlete.

APPARATUS FOR MUSCULAR WORK.

Of particular interest is the ergometer employed in these experiments. Two instruments were actually used, ergometer I being the identical instrument employed in the previous research by Benedict and Carpenter," and ergometer II a replica of the same with slight structural modifications but with no alteration in principle. Ergometer I was used in all experiments up to January 19, 1912, inclusive, the experiments subsequent to that date being made with ergometer II. Both instruments have previously been described and the details given of a series of calibrations made with them by Benedict and Cady."

In the latest form of bicycle ergometer (see ergometer II in the frontispiece), a bicycle frame, sprocket, and pedals were used, the rear wheel being replaced by a large copper disk, 405 millimeters in diameter and approximately 6 millimeters thick, securely fastened to a ball-bearing hub. An electromagnet (shown in the frontispiece) was attached to the bicycle frame by means of brass piping, and the disk so adjusted as to rotate exactly in the middle of the electro-magnetic field between the pole faces, with an airspace on each side of the disk of approximately 1 millimeter. The upper edges of the pole-faces form a line that is tangential to the circumference of the copper disk. The large eddy-currents generated as the disk passes through the magnetic field are short-circuited in the disk; thus the whole instrument has the effect of an electric brake, heat being developed in the copper disk

a Benedict and Carpenter, U. S. Dept. Agr., Office Expt. Stas. Bul. 208, 1909.

Benedict and Cady, Carnegie Institution of Washington Publication No. 167, 1912; also, Cady and Benedict, Physikal. Zeitschr., 1912, 13, p.920.

and rapidly radiating to the surrounding air. By varying the intensity of the magnetic field, a greater or less brake-effect can be produced. In ergometer II the resistance of the magnet is 10 ohms, and a current of 1.5 amperes through the coil produces substantially the same brake-effect as 1.25 amperes on ergometer I. The heat developed in the magnet of ergometer II by magnetization with a current of 1.5 amperes is 17.8 calories per hour.

With the excessively high rates of speed frequently employed by the subject, it was of course impossible to count the ergometer revolutions from minute to minute, and hence a counter of a standard make was attached. Two of these were employed to insure accurate records, one actuated by direct contact with the pedals and the other by a pin on the large sprocketwheel.

The ergometer was calibrated by using one of the respiration calorimeters (the chair calorimeter ") in the Nutrition Laboratory. The instrument was first divested of the handle-bars and pedals, then placed inside the respiration chamber, and fitted with a flexible shaft which could be driven from the outside by an electric motor. By exciting the magnet and rotating the disk, heat was produced which could be directly measured by the calorimeter as in an ordinary calorimeter experiment. Different strengths of current through the magnetic field and varying speeds were used in calibrating the two instruments, a large number of calibration curves being obtained for both ergometers. With these curves it is necessary to know only the number of revolutions and the intensity of magnetization in order to compute directly and accurately the muscular work which is transformed into heat.

Since the greater number of experiments reported in this book were made with ergometer II, a few characteristic calibration curves of this instrument are given in fig. 1, i. e., those for currents through the magnet of 0.5, 0.95, 1.10, 1.25, 1.35, and 1.5 amperes, the six excitations of the magnet used. All of the curves for the calibration of both ergometers I and II, with the exception of that for 0.5 ampere, are given in detail in an earlier publication."

The ergometer is very accurately constructed and runs without much friction. By again employing the respiration calorimeter, friction tests were made with both ergometers I and II. Two tests with ergometer I gave as results 0.000274 and 0.000157 calorie per revolution respectively; three tests with ergometer II gave 0.000355, 0.000182, and 0.000351 calorie per revolution. While the calorimeter used for the tests was primarily designed to measure the heat production of a man and hence was not as well adapted for the measurement of so small an amount as 1 or 2 calories per hour, the values obtained for the two ergometers agree reasonably well and are doubtless not far from correct. The details have already been published."

Inasmuch as the question of the internal friction of the legs came into discussion, it seemed desirable to rotate the ergometer by means of an electric motor so that the legs of the subject would freely move up and down as the pedals revolved. For this purpose a split wooden pulley, with a groove around the periphery, was attached to the hub of the copper disk and connected with a belt brought out to the pulley on the armature shaft of the motor. By measuring the intensity of the field, the current through the armature

a Benedict and Carpenter, Carnegie Institution of Washington Publication No. 123, 1910.
b Benedict and Cady, loc. cit., pp. 18-28.

c Ibid., pp. 21 and 29.

.023

.022

.021

.020

1.50 AMP

of the motor, and the voltage between these points, the amount of work required to rotate the machine could be computed. It was hoped by this means to devise some simple and rapid method for calibrating the machine without using the more expensive method of determining the heat produced by the respiration calorimeter. Owing to the great loss by friction and the tension of the belt, the preliminary experimenting by this method was very unsatisfactory, and time did not permit further study along this line. The electric motor was, however, extensively used to drive the ergometer in the so-called "motordriven experiments."

1.35 AMP

.019

125 AMP

.018

.017

LIO AMR

.016

.015

95 AMR

.014

.013

.012

.011

.010

It is important to note that the ergometer, when once calibrated, gives directly the number of calories of external muscular work put upon the machine for each revolution of the pedals, the calories varying somewhat with the rate of speed of revolution. Fortunately for the speeds most commonly used by bicycle riders, namely, between 60 and 80 revolutions of the pedals per minute, when the magnetizing current is constant, the energy output per revolution is practically constant. At speeds lower or higher than this, there is a marked decrease in the heat per revolution. The peculiar anomalies of these calibration curves have been the subject of special discussion elsewhere."

.004

30 40 50 60 70 80 90 100 110 120 130

FIG. 1. - Calibration curves showing heat per revolution of ergometer II. The revolutions per minute are given at the bottom of the figure, and the calories per revolution at the left-hand margin.

APPARATUS FOR DETERMINING THE GASEOUS EXCHANGE.

A respiration apparatus that would at one and the same time provide sufficient ventilation for a man working almost to the limit of human endurance and permit both the complete absorption of carbon dioxide and an accurate measurement of oxygen was difficult to obtain, but by slightly modifying the form of respiration apparatus employed in this laboratory for a number of years we were able to meet the requirements for experiments with muscular work. This apparatus has recently been described in detail, including the modifications necessary for experiments with muscular work," together with a diagram showing the particular adjustment of the apparatus for such experiments. Since the description was published, however, further modifications were found desirable in the progress of this research.

The connections of the respiration apparatus as used for muscular-work experiments are shown in fig. 2. As may be seen in this figure, a rotating ventilator or blower draws air along a pipe and forces it through two Woulff

a Benedict and Cady, loc. cit.; Cady and Benedict, loc. cit.
b Benedict, Deutsch. Archiv. f. klin. Med., 1912, 107, p. 156.

bottles containing sulphuric acid. The greater part of the water-vapor in the air-current is absorbed by the acid in the first bottle, but any traces which

a

may remain are removed by the acid in the second bottle. The dry air is then passed through the carbon-dioxide absorbers, i. e., two bottles containing sodalime. Since the air absorbs water from the soda-lime it is necessary to remove this moisture by passing the air-current again through sulphuric acid. When the air leaves the drying-bottle it is deficient in oxygen and

free from carbon dioxide and water-vapor, but inasmuch as perfectly dry air, if inspired by the subject, would absorb moisture from the respiratory tract and produce discomfort, moisture must be added to the air before it is returned to the subject for breathing. This is done by passing the air-current through the lower part of a Kipp generator partially filled with water. A small amount of sodium carbonate is added to the water in this vessel, which effectually absorbs any acid fumes that may have been carried over. The deficiency in the oxygen content is made up by the addition of a measured amount of the gas at O2. The air then passes through a 3-way valve and returns, either directly or through the short circuit connected with the subject, to the tension-equalizer and the blower. By weighing the carbondioxide absorbers, and the air-drying vessel before and after the experimental period, and noting the amount of oxygen consumed either by finding the loss in weight of the cylinder of oxygen, or by measuring with a carefully calibrated gas-meter the volume of oxygen added to the air-circuit, the carbon-dioxide production, the oxygen consumption, and the respiratory quotient may be computed.

The absorbing system as here described and used in the muscular-work experiments differs from the usual adjustment of the apparatus in several particulars. When the subject is engaged in severe muscular work, an excessive amount of carbon dioxide is produced, sometimes amounting to as much as 2,500 c.c. per minute. To provide for the absorption of this large amount of carbon dioxide, it was considered safer to use two soda-lime bottles connected in series, instead of one carbon-dioxide absorber as usual. In the most recent form of the respiration apparatus a spirometer is used, and much emphasis is placed upon the character of the respiration as a As a matter of fact, to guard against oxygen-want in the apparatus, the air in the closed system was always arbitrarily adjusted somewhat oxygen-rich. Benedict, Deutsch. Archiv. f. klin. Med., 1912, 107, p. 172.

b