A STUDY OF OXYGEN METABOLISM. 449 



however, this correlation does not manifest itself when curves 

 for temperature and pupal weights are compared. Moreover, 

 a constant temperature serves merely to limit somewhat the 

 range of fluctuation, instead of insuring stability, as can be seen 

 from Table III. Temperature, then, operates only in a large 

 way in its influence on weight of pupae. 



Corresponding to the sudden drop in metabolism shortly after 

 the inauguration of pupal life, there occurs a marked drop in 

 pupal weight. Sometimes this loss is relatively enormous, in 

 other exceptional cases relatively insignificant. Unlike metabo- 

 lism, however, there is no recovery in weight after the second 

 day, the level for the third and fourth days remaining approxi- 

 mately identical with that for the second day. Sometimes for 

 these latter days a slight increase is registered but as frequently 

 a slight fall occurs. 



The interesting sequence of changes during the pupal period 

 in respect to pupal weights and metabolism is presented in Fig. 3. 

 The data on which this figure is based represent a mating selected 

 for illustration because a large number of determinations is 

 available and the pupal weights show an unusual consistency of 

 level. The figure is divided into parts corresponding to the 

 several days of pupal life and a common scale is used which 

 enables one to carry out comparisons between any two days. 



The relationship between the curves for metabolism per pupa 

 and per unit body weight affords an index of the ratio existing 

 between weight of respiring tissue and that of inert materials 

 associated with the respiring tissue. When the two curves follow 

 a strictly parallel course this ratio must remain constant in the 

 several lots of pupa?, provided complicating factors are not 

 present. When for a particular determination the curves become 

 approximated, the inference may be made that the relative 

 amount of non-respiring or inert materials is reduced, whereas, 

 when they become more widely separated, this condition may 

 be attributed to a relative increase in the non-respiring materials 

 present. It will be clear, therefore, that in general the ratio of 

 respiring to non-respiring substances in the pupse on which Fig. 3 

 is established remains a fixed quantity. Interesting exceptions, 

 however, appear in the first and last determinations, numbers 8 



