and it is the need for the production of heat that regulates the total demand for Calories which every organism requires from its ration. It is not because it produces too much heat that the organism gets rid of it peripherally: it is rather because it inevitably disperses it that it is adapted to produce it.
Rubner's Experiments.—This conception of the rôle of alimentation is based on two arguments. The first is furnished by Rubner's last experiment (1893). A dog in a calorimeter is kept alive for a rather long period (two to twelve days); the quantity of heat produced in this lapse of time is measured, and it is compared with the heat afforded by the food. In all cases the agreement is remarkable. But is it possible that there should be no such agreement? Clearly no, because there is a well-known regulating mechanism which always exactly proportions the losses and the gains of heat to the necessity of maintaining the fixed internal temperature. This first argument is, therefore, not conclusive.
The second argument is drawn from what has been called the law of surfaces, clearly perceived by Regnault and Reiset in their celebrated memoir in 1849, formulated by Rubner in 1884, and beautifully demonstrated by Ch. Richet. In comparing the maintenance rations for subjects of very different weights, placed under very different conditions, it is found that the food always introduces the same number of Calories for the same extent of skin—i.e., for the same cooling surface. The numerical data collected by E. Voit show that, under identical conditions, warm-blooded animals daily expend the same quantity of heat per unit of surface—namely, 1.036 Calories per square yard. The average ration intro-
