INSECT ECOLOGY 357 



temperature to secure a thermal or physiological constant cannot be 

 based on a mere addition where variable temperatures are involved, for 

 it is evident that every degree has a different value in relation to the 

 time factor. Thus as the mean temperature rises with the advance of 

 season both the time for the pupal stage and the total accumulated 

 .perature for the pupal stage of the codling moth decrease with the 



vancing season. Though a fairly constant ' total effective tem- 



irature' for any given phase of an insect's life or activity may be 

 ,ecured for the summer months when there is a fairly constant mean 

 ;emperature, such an accumulation will have no meaning in regard 

 ;o the same phenomena in spring and fall when the temperatures are 



tore variable. If we wish to be exact, we must secure the temperature 

 curve for the species, based on the observation of a considerable num- 

 ber of individuals kept at different constant temperatures, or possibly 

 better at temperatures having a diurnal variation with constant maxi- 

 mum and minimum, and with fairly constant moisture conditions." 

 Krogh on Temperature -velocity . The results obtained by Krogh, 



r hich differ in some respects from those of other investigators, are 

 egarded as highly important. He finds that the temperature-velocity 

 ve expressing the rate at which segmentation takes place in frog's 



;gs is, between 7 and 20.7, a straight line. "An increase in tem- 



:rature between these limits produces a proportional increase in the 

 elocity with which the processes in the egg leading up to segmentation 

 :e place. Below 7 the curve deviates from the straight line and the 

 eaction takes place more rapidly than one would expect from the 

 esults obtained at higher temperatures. At the lowest temperature, 



here the development certainly is no longer normal the curve turns 



ownward once more." 



"The relation between the temperatures and the velocity of embry- 

 onic development is algebraic over a range of temperatures which 

 corresponds approximately to that at which normal development can 

 take place, and the curve representing the relation is consequently a 

 straight line." The velocity of embryonic development is a linear 

 function of the temperature. 



In regard to the relation between temperature and the later stages of 

 development of the frog, Krogh says: "Between the temperatures 12 

 and 25 the increment in velocity of the embryonic development of the 

 frog is therefore proportional to the temperature increment, but below 

 12 the development is more rapid than one would expect from the 

 formula." 





