376 HARVEY M. SMITH. 



of the same individual, and a variation in the average size from 

 different individuals of the same species. 



5. A number of factors tend to complicate the results on carbon 

 dioxide production. The most important is body weight, the 

 smaller animals producing more carbon dioxide per unit of weight 

 than the larger animals. This factor becomes especially impor- 

 tant in making comparisons of different species, because species 

 vary quite widely in their average w r eight. 



Starvation in Necturus seems to cause an increase in the carbon 

 dioxide output per unit of weight, even when all results on an 

 individual are reduced to a constant weight value. In the other 

 species studied starvation seems to lead to a decrease in carbon 

 dioxide production. 



Other factors complicating carbon dioxide output results are 

 daily variations in the metabolism of the individual; sex, the 

 male producing slightly more than the female; disease, usually 

 resulting in an increased production, a fact which suggests the 

 fever response in man; and the movement of the animal. This 

 last factor may become of extreme importance if the animal is 

 active, but an attempt has been made to eliminate such results 

 by keeping the animals quiet and discarding results of deter- 

 minations made on actively moving animals. 



6. The conclusion that cell size varies inversely with meta- 

 bolic activity is justified by the evidence presented. This is 

 shown in a general way by comparing classes of Vertebrata, and 

 more specifically by detailed results on twelve species of 

 Amphibia. 



The writer wishes to express his gratitude to Dr. M. F. Guyer, 

 under whose direction the work was undertaken, and to Dr. A. S. 

 Pearse and Dr. H. C. Bradley for their many helpful suggestions 

 and criticisms. 



