в period queen larvae. 13 fully representative larvae (\vith weights from 

 7-1 to 11-3 mg and wing areas from 40 to over 70 x 10-^ sq. mm) had identical 

 brain cells (Fig. lb). No purple granules were present at all: the cells were 

 larger, the nucleolus more conspicuous, and the cytoplasm grey and finely 

 granular. They appeared to be active but differently so. All the cc were also 

 devoid of purple granules. 



В period worker larvae. 13 representative larvae (with weights from 2-0 

 to 4-4 mg and wing areas from 2-5 to 10 x 10-^ sq. mm) all had brain cells 

 quite different from the queen larvae: they were of the size characteristic 

 of the A period but contained fewer purple granules (Fig. Ic). The cc had 

 granules in 3 and none in 10 of the Î3 cases. Thus the impression is given 

 of a waning A period type of activity. 



A period autumn larvae. 12 representative larvae collected in October 

 including both small pre-diapause and large diapause ones all showed brain 

 cells that were no larger than normal nerve cells and had no purple granules 

 at all. The cc also had no purple granules. Five large diapause larvae examined 

 at the end of November were identical and five after a period of six day's 

 incubation were also in this state. Evidently one of the things that low winter 

 temperatures do, is to reactivate the brain cells. 



Corpora allata 



These glands have been examined in the same larvae. In A period spring 

 larvae they range in diameter from 40—65 [x with a mean of 55-5 and the cells 

 have large nuclei containing black chromatin and grey vacuolated cytoplasm 

 (Fig. le). In В period queen larvae they are much larger with diameters rang- 

 ing from 65 to 90 [jL and averaging 78-0 (Fig. If). This is due to the greater 

 size of both nuclei and cytoplasm; the latter also has fewer and smaller vacuo- 

 les. In В period worker larvae, on the other hand, the ca are not much larger 

 than in the A period and range from 50 to 70 уь averaging 61-0. Analysis 

 of variance shows that the difference between queens and others is highly 

 significant but that between В period worker larvae and A period larvae 

 is not so. 



The В period queen larvae are of course larger than the worker larvae 

 of that period: their mean weight is 2-5 times that of the workers. It is reason- 

 able to assume that the volume of the ca is proportional to its diameter cubed, 

 and it is then easy to show that the average for queens is 2-1 times that for 

 workers and hence the ca are not by any means greater relative to body size 

 in queen larvae and may in fact be considered allowing for various inaccur- 

 acies, to be about the same. 



The autumn В period larvae are interesting for whereas October ones are 

 normal, in late November there is striking evidence of change (Fig. Ig). The 



169 



