Further analysis showed Rubner2 that this evenness of heat production per unit of body surface was not due to any relation between the area of body surface and the area of cell surface within the organism.

Rubner estimates that a man weighing 60 kilograms contains 37.8 kilograms of cell mass, of which 40 per cent, is in muscle tissue, and that while the absorptive surface of the intestinal tract is 1.5 square meters, the surface area of the body cells amounts to 9014 square meters (2.2 acres). There are in 1 kilogram of body weight of man 150.2 square meters of such surface, and each square meter of cell surface produces at least 0.2 calorie per day. In the newborn mouse each square meter of cell surface produces eleven times this amount, or 2.2 calories. It is of interest, also, to note that a kilogram of yeast cells presents a surface area of 600 square meters and at a temperature of 380 C, or that at which mammalian cells exist, 1.25 calories per square meter of surface are produced in twenty-four hours, 8.34 grams of cane-sugar undergoing inversion and fermentation during that interval.

1 Rubner: "Energiegesetze," 1902, p. 174.

2 Rubner: "Archiv fur Physiologie," 1913, p. 240.

This reaction is independent of the strength of the sugar solution within the wide limits of 2.5 to 20 per cent. If the strength of the solution be at the maximum of normal reaction, or 20 per cent., the quantity of sugar utilized in twenty-four hours would be contained in a film 4/100 millimeter in thickness surrounding the cells. A like analysis shows that in man, whose cells are bathed in a medium containing 0.1 per cent, of sugar, the quantity necessary for the support of life during one day would be contained in a layer which if spread around the cell would be 5/100 millimeter in thickness.

From the calculation of the energy requirement in the food for the life of a man to the energy liberated by a yeast cell in its simple resolution of sugar into alcohol and carbon dioxid is indeed a far cry, except as showing that the energy doctrine, as enunciated by Rubner, unites the world of living things.

Magnus-Levy1 made 41 short time respiration experiments on the same man when resting without food. The greatest variations from the mean were - 7 and + 10 per cent. The calories per square meter per twenty-four hours were 812. In 1912 Lusk2 calculated that the heat production of three quiet and sleeping dogs was 759, 748, and 746 calories per square meter of surface at an environmental temperature of 260 C, that a dwarf produced 775 calories per square meter of surface, and that four out of five sleeping men investigated by Benedict showed an average heat production of 789 calories per unit of area. Only in the sleeping infant 7 months old, investigated by Howland, did the metabolism appear out of the ordinary and reached a level of 1100 calories, and this factor was specifically pointed out as indicating a higher metabolism in the youthful protoplasm than is present in the adult.

The critical studies of F. G. Benedict3 led him to conclude "that the metabolism or heat output of the human body, even at rest, does not depend on Newton's law of cooling,1 and is, therefore, not proportional to the body surface".

1 Magnus-Levy: "Pfluger's Archiv," 1894, lv, 1.

2 Lusk (with McCrudden): "Journal of Biological Chemistry," 1913, xiii, 450.

3 Benedict, F. G.: "Journal of Biological Chemistry," 1915, xx, 298.

That a greater metabolism is induced in man after the ingestion of a liter of cold milk than after taking the same amount when it is warm, was shown in Tangl's2 laboratory, and indicates that an influence may be exerted by cooling. The body temperature fell 0.25 to 0.8 degree. That such an influence is exerted by cooling was clearly demonstrated by Lusk,3 who compared the heat production after giving glucose dissolved in cold water and in water at the body temperature to a dog placed in a calorimeter, with the following results:

* Plus heat for warming the cold water.

When warm water was ingested the computed heat production agreed with that actually found, but when cold water was given there was an increased oxidation, as shown by indirect calorimetry, in order to provide for the body heat lost to the fluid in the stomach (see p. 132).

Benedict is in agreement with Carl Voit when he concludes that the mass of active protolasmic tissue determines the height of the metabolism. However, in the search for a standard upon which to calculate what would be the normal heat production of a man suffering from disease it is obviously impossible to measure the mass of active protoplasmic tissue.

1 This law reads, "The quantity of heat gained or lost by a body in a second is proportional to the difference between its temperature and that of the surrounding medium." At the higher temperatures of environment it is obvious this law does not control. See Rubner's experiments on guinea-pigs, p. 120, demonstrating that the effect of cold on the skin is not a sufficient explanation of the law of skin area.

2 Hari and von Pesthy: "Biochemische Zeitschrift," 1912, xliv, 6.

3 Lusk: "Journal of Biological Chemistry," 1915, xx, 578.

When the Russell Sage Institute of Pathology constructed in Bellevue Hospital an Atwater-Rosa calorimeter copied in the main after the successful models of Benedict, it became absolutely essential that some criterion of normal metabolism be established as a basis from which one could estimate whether the metabolism of a patient under investigation was higher or lower than the normal. The severe criticisms of Benedict upon the method of estimating heat production from the unit of surface led to a very careful review of all the evidence and to new experiments. Du Bois,1 who took up this work, has used an accurate and ingenious method with which he has been able actually to measure the surface area of normal men. He covered the body surface with tight-fitting underwear, applied melted paraffin, and then paper strips to prevent change in area when the covering was removed. This model of the surface when cut into flat pieces was photographed upon paper in which equal areas were of equal weight. From the weight of paper which received the photographic impression the area of body surface could readily be calculated. A round ball having an area of 0.1490 square meter, when measured by this method, gave an area of 0.1488 square meter. After this fashion E. F. and Delafield Du Bois have discovered that the formula heretofore used for estimating the surface area in man showed an average inaccuracy of 16 per cent, and a maximal variation from the normal of 36 per cent., this being found in very fat individuals. Two new formulae, a "linear" and a "height-weight" formula, have been evolved which give an average variation of ± 1.5 per cent, and a maximal variation of ± 5 per cent. Using the older formula of Meeh, the heat production per square meter of surface is 833 calories during twenty-four hours, but using the more accurate formula of Du Bois this rises 16 per cent, to 953 calories. In normal adults of various shapes and sizes this is the basal metabolism as measured when the individual is resting and before the administration of food in the morning.

1 Du Bois, D. and E. F.: "Archives of Internal Medicine," 1015, xv, 868; Ibid., 1016, xvii, 863.