72 PRINCIPLES OF ANIMAL NUTRITION. 
tion apparatus of very large dimensions. Recently Atwater & 
Rosa * have constructed a form of Pettenkofer apparatus for use 
as an animal calorimeter in which the method of measuring and 
sampling the air current has been materially improved and rendered 
more accurate. 
When the Pettenkofer apparatus is employed for experiments 
upon herbivora, special provision is necessary for the determination 
of the gaseous hydrocarbons excreted in considerable quantities 
by these animals. This is accomplished by passing a sample of the 
air coming from the apparatus through a combustion-tube contain- 
ing copper oxide, or preferably spongy platinum (platinized kaolin), 
heated to redness. The hydrocarbons are thus oxidized and the 
resulting carbon dioxide determined. 
Pettenkofer & Voit,t in their earlier investigations, deter- 
mined the excretion of combustible gases by a dog, with the follow- 
ing results per day: 
Food. 
Hydrogen, Methane, Carbon 
Meat, Fat, Starch, Grams, Grams. eg 
Grams. Grams. Grams. 
B00 | sisiescenaainys 200 7.2 4.1 416.0 
B00 |aisavscnscacs 200 5.2 6.3 420.6 
500 sEelanesiek ieee 200 7.2 4.7 428.2 
500 200 seleigevsatcieivies 6.4 3.7 417.3 
500 200° leeicss seeds 4.3 4.5 427.8 
According to the above figures, a trifle less than 3 per cent., 
on the average, of the total carbon excretion was in the form of 
methane. No similar determinations seem to have been made by 
Pettenkofer & Voit in their later experiments, and it appears to 
be generally assumed that they are unnecessary in investigations 
upon man and the carnivora. 
The Zuntz Apparatus.—Both the Regnault and the Pettenkofer 
types of apparatus are calculated for the determination of the 
total gaseous excreta of lungs, skin, and digestive tract through 
considerable periods of time, and their use enables us to compare 
the total income and outgo of carbon. 
*U. 8. Dept. Agr., Office of Experiment Stations, Bulletins 44 and 68. 
+ Ann, Chem. Pharm., Supp. Bd. II, p. 66. 
