446 



M. R. CLARE. 



of the cultures. This condition resulted in the production of a 

 limited number of very small pupae of exceptionally low weights. 

 A single example from this series has been retained in Table 5 

 in order to show that, in spite of the very low pupal weights, 

 the rate of metabolism is to all appearances normal. The rather 

 drastic elimination which has been practiced has reduced to a 

 considerable degree both the number of determinations and the 

 variability in rates, but that the latter has been by no means 

 removed is shown in Fig. 2. 



26 . 



01 



a. 22 



t. 18 



01 



o_ 



Q H . 



6 



a 10 

 u 



Hating Fi 



V 



FIG. 2. Metabolism curves showing variability in rates within a single mating, 

 cultured at 25 C. Numbers indicate particular manometers used in establishing 

 rates. Ordinates = Oz per gram per minute, in cu. mm. 



There are several conceivable factors which might underlie 

 this variability and they will be taken up in turn. It is obvious 

 that if comparisons are to be made between the several stocks in 

 respect to oxygen consumption this irregularity must be reduced 

 to a minimum. Suspicion was at once cast on the calibration 

 of the manometers used in making the determinations. That 

 these are not the primary source of the variability will be evident 

 from a further consideration of Fig. 2. The curves of this figure 

 are for lots of pupae all of which are offspring of a single pair of 

 flies. Associated with each curve is a number which is that of 

 the manometer used in determining the rates. For each of two 

 manometers, numbers 6 and 9, three sets of determinations are 

 represented which, it should be noted, exhibit marked variability 

 in rates of metabolism. The manometers were very carefully 



