THE FOOD SUPPLY 499 



sions are to pass beyond the empirical stage and lead to the establish- 

 ment of general principles. For this purpose Kellner has used a form 

 of the so-called Pettenkofer respiration apparatus, first devised many 

 years ago by Professor v. Pettenkofer in Munich, while the Pennsyl- 

 vania institution employs an instrument known as the respiration calo- 

 rimeter, first devised by the late Professor W. 0. Atwater for investi- 

 gations in human nutrition and which has been enlarged and modified 

 to adapt it for experiments upon domestic animals. 



The central feature of both apparatuses consists of an air-tight 

 chamber through which a measured current of pure air passes and 

 within which the animal stands in a comfortable stall, where it can be 

 fed and watered at will. The total energy contained in the feed of the 

 animal is ascertained by determining the amount of heat which a sample 

 of it produces when completely burned, while the energy escaping in the 

 visible excreta is measured in the same way. Furthermore, by analyz- 

 ing samples of the air-current before and after its passage through the 

 chamber containing the animal the gaseous waste products given off 

 by the latter are determined. 



Finally, energy escapes from the animal in the form of heat. In 

 the German experiments the amount of heat produced by the animal 

 is virtually computed from the amount and kind of materials oxidized 

 in the body. This may also be done in the experiments with the respira- 

 tion calorimeter, but in addition this apparatus is provided with appli- 

 ances for the direct determination of the heat given off, it being taken 

 up by a current of cold water circulating through copper pipes and its 

 amount measured with the aid of sensitive thermometers. In this way 

 the total income and outgo of the animal can be compared, the differ- 

 ence showing how much of the energy of the food has been stored up as 

 meat or fat, while a comparison of the observed with the computed heat 

 production serves as a check on the accuracy of the experiments. 



The method is not unlike that employed in locomotive testing plants 

 like, e. g., that of the Pennsylvania Eailroad at Altoona. Just as in 

 the latter, the heat value of the fuel is measured, so in the experiments 

 upon the animal the heat value of the feed, which is the fuel of the 

 animal body, is determined. The losses in the visible excreta of the 

 animal may be compared to unburned coal dropping through the grate, 

 while the gaseous excreta correspond to the flue gases. A large amount 

 of heat is given off in both cases, and the final balance of income and 

 outgo makes it possible to trace exactly the use which the locomotive or 

 the animal makes of the energy supplied to it. 



The material or ration to be tested is fed for some three or four 

 weeks with the greatest regularity. During the latter portion of this 

 time, after the effect of the ration has become fully established, the 

 animal spends from two to five days in the respiration apparatus or the 



