Recent experiments by Moulton1 show that the nitrogen content of cattle is almost exactly proportional to the surface area of the animal. If the nitrogen content be a measure of protoplasmic tissue, these experiments afford a striking confirmation of the doctrine of Voit that the heat production is proportional to the mass of living cells (see p. 45).

However, the remarkable experiments of Grafe and Graham2 appear to indicate that notwithstanding a large addition of protein to the organism of a dog following fasting, there is no increase in the fundamental metabolism. These experiments are summarized below:

After seven days of food the animal recovered all the weight and nitrogen lost during twenty-one days of fasting and its heat production was as before the fast. (See also p. 211).

The organism, therefore, preserves the tropical temperature of its cells at the expense of a metabolism which is proportional to the skin area of the individual.

1 Moulton: "Journal of Biological Chemistry," 1916, xxiv, 299.

2 Grafe and Graham: "Zeitschrift fur physiologische Chemie," 1911, Ixxiii. 1.

The loss of heat by an organism at rest follows these paths:

1. Conduction and radiation.

2. Evaporation of water from lungs and skin.

3. Warming the food ingested.

4. Warming the inspired air (conduction).

The great outlets for heat loss are by conduction and radiation (of which in the dog 97.3 per cent, takes place through the skin and 2.7 per cent, through the lungs1) and through the evaporation of water. The losses through warming the food, and through heat of the urine and of solution of urinary constituents through the feces, and the warming of expired carbon dioxid may be ordinarily disregarded.

The pathway for the loss of heat varies with the temperature of the environment. At a low temperature there is little evaporation of water, and at a temperature of 370 C. there can be no heat loss by radiation and conduction (except by a rise in body temperature), and water evaporation removes the whole of it. In the dog at a high temperature there is spreading out of the limbs to promote heat loss by radiation and conduction, and rapid breathing (polypnea) with extension of the hyperemic tongue to promote evaporation of water. In the horse and in man there is especially an outbreak of sweat which is not possible in the dog, as its skin does not secrete sweat.

Du Bois2 finds that the average loss of water from the lungs and skin is 680 grams per day in the normal resting man at an environmental temperature of 230 C. and medium humidity. The evaporation of this amount of water represents an absorption of heat equal to 24 per cent, of the total heat loss. This latter figure is in exact agreement with an average of results previously reported by Benedict and Carpenter.3 According to Loewy4 the loss of water of perspiration per square meter of body surface is greatest in the arms, next greatest in the legs (the extremities yielding not far from 75 per cent, of the total), and least from the trunk. The greatest actual loss is, however, from the legs.

1 Rubner: "Energiegesetze," 1902, p. 187.

2 Gephart and Du Bois: "Archives of Internal Medicine," 1916, xvii, 902.

3 Benedict, F. G., and Carpenter: Carnegie Institution of Washington, Publication 126, 1910.

4 Loewy: "Biochemische Zeitschrift," 1914, lxvii, 243.

Loewy also finds that in men without sweat-glands the evaporation of water from the skin may amount in maximo to 15.6 grams per square meter per hour, or 800 grams for the whole body during a day. Vasomotor reflexes may play an important part in the quantity of water evaporated. Placing the right forearm in cold water reduced the water elimination from the right leg from 3.64 to 3.22 grams per square meter of surface per hour. Washing the right arm with alcohol and ether reduced the water elimination of the right leg to 1.78 grams for the same unit of measurement. In both of these experiments the leg showed an increase above the normal evaporation of water after the removal of the stimulus of cold from the arm.

Generally speaking, there is little difference between the temperature of the inner organs of the body. Heidenhain,1 confirming earlier work of Claude Bernard, found that in 84 out of 94 experiments with dogs the temperature of the right ventricle was higher than that of the left, two-thirds of the cases showing differences between 0.1° to 0.30. Claude Bernard2 states that during digestion the blood of the hepatic vein is 0.1° higher than that of the portal vein. Quincke8 found that the temperature of the empty stomach of a boy was constantly 0.120 higher than the rectal temperature, and that after the ingestion of 500 c.c. of water at a temperature of 200 C. the original temperature was not regained for seventy to seventy-five minutes. Rancken and Tigerstedt4 find a temperature in the stomach of a boy with a gastric fistula which averages 0.09° higher and is in maximo 0.20 higher than that of the rectum.

1 Heidenhain: "Pfluger's Archiv," 1871, iv, 558.

2 Bernard: "Lemons de physiologie operatoire," Paris, 1879, p. 481.

3 Quincke: "Archiv fur exp. Path, und Pharm.," 1889, xxv, 375.

4 Rancken and Tigerstedt: "Skan. Archiv fur Physiologie," 1909, xxi, 85.

Regarding the surface temperature, Henriques and Hansen1 report the following temperatures at different depths in the fat of the hog's back just one side of the median line:

The environmental temperature was, unfortunately, not noted. It must be evident that under these conditions blood coming from the internal organs must lose heat to the cooler surface of the organism.

Benedict and Slack2 studied the simultaneous records of rectum, vagina, axilla, breast, groin, hand, arm, and mouth, and concluded that aside from the skin temperature a rise or fall in rectal temperature is accompanied by a corresponding rise and fall in temperature of all other parts of the body in man.

Coleman and Du Bois3 find that in fever, under conditions of a changing blood-supply to the skin, well-covered surface thermometers give a more accurate indication of the average change in body temperature than does the rectal thermometer. As the measurement of the amount of heat gained or lost by an organism during an experiment in which direct calorimetry is determined is effected through the observation of the changes of body temperature, this is an important matter. It may be stated as a general principle that when there is a wide variation in rectal temperature direct and indirect calorimetry do not usually agree as closely as when there is little alteration in body temperature, which indicates that the blood is not at all times so distributed throughout the body that the average rise throughout all the parts is equal to the rise in the recturn alone. Yet, on the whole, the rectal temperature is the best guide available.

1 Henriques and Hansen: "Skan. Archiv fur Physiologie," 1901, xi, 161.

2 Benedict and Slack: Carnegie Institution of Washington, Publication 155, 19" 3 Coleman and Du Bois: "Archives of Internal Medicine," 1915, xv, 887.