A STUDY OF OXYGEN METABOLISM. 447 



calibrated several times and any possible errors from this source, 

 although carefully looked for, have eluded detection. 



PUPAL WEIGHTS AND METABOLISM. 



In studies on respiratory metabolism the practise has been to 

 emphasize the relation between weights of experimental organ- 

 isms and intake of oxygen or production of carbon dioxide, on 

 the assumption that for comparable samples of living material 

 within a species, the rates of metabolism remain fairly constant. 

 Rubner, along with others, on the other hand, has instituted the 

 procedure of basing metabolism on extent of body surface, 

 arguing, in the words of Krogh (1916) that "the metabolism is 

 simply a function of the conditions for loss of heat, while there 

 is no such thing as a specific oxidative activity of the cell." 

 Rubner later saw fit, however, to qualify this idea. These 

 practises were derived from investigations on warm-blooded 

 animals possessed of a heat-regulating mechanism, and their 

 applicability to invertebrate animals is extremely uncertain. 

 Unfortunately, the studies on metabolism in invertebrates have 

 been so few in number and have afforded results of so conflicting 

 a character that conclusions based on them do not seem to be 

 justified. We feel obliged, therefore, to examine the data for 

 Drosophila with some care and determine, if possible, the signifi- 

 cance to be attached to weight; and attempt to decide whether 

 or not rates of metabolism are subject to decided change. No 

 data are available on which to base figures for pupal surface. 



Pupal weights exhibit a \vide range in value, the means for 

 newly-formed or first-day pupae in lots of 10 varying between n 

 and 15 milligrams. At the extremes of the range lots of 10 

 pupae may possess a weight as low as 9 mg. or as high as 17 mg. 

 For the products of a single mating, likewise, the range of 

 variability is very pronounced. 



In addition to the roles played by food and overcrowding as 

 factors affecting pupal weight and these can hardly be con- 

 sidered as applying in this work temperature certainly is a 

 determining agency. An examination of Table 2 shows that 

 pupal weights of the second period are significantly lower than 

 those for the first period. In the summer of 1923, during a 



