STUDIES ON INSECT SPERMATOGENESIS. 413 



usual course of events and in the characteristic stages of large sperm 

 formation, the appearances are normal for the large sperm. It is 

 possible that this exception was due to a disturbance in the customary 

 amount of difference between the large and small generations of sper- 

 matocytes, of which there were some slight indications. 



Conclusion. 



In the preceding section the cause and meaning of polymegaly have 

 not been considered, because thus far I have been unable to obtain 

 any evidence bearing on either of these points. Montgomery first 

 ('98) concluded that the larger spermatocytes were "due to their 

 receiving a greater amount of food, . . . That is to say," they "must 

 be nourished by a richer blood supply." Subsequently ('10), he 

 modified the latter statement, stating that the nutritional diflFerences 

 "are due to differences in the follicular nurse cells of the testis." 

 This explanation really explains nothing, for we are still in the dark 

 as to the reasons why certain follicular cells should become larger than 

 their neighbors. Indeed, it is not e^^dent why the enlargement of the 

 nurse cells may not as well be a result as the cause of the large genera- 

 tions of spermatocytes. In any event, the morphological condition 

 of the nurse cells helps us not at all to understand the actuating 

 mechanism behind the whole phenomenon. Furthermore, the un- 

 usual case cited in the preceding section of an individual in which 

 cysts of cells normal in size occur together with cysts of large cells 

 in the same lobe, would seem to indicate that nutritional differences 

 are not the only factor at work. 



It may be noted here that other cases of polymorphism are known 

 to occur in the male germ cells of several animals, and such differences 

 may be of more general occurrence than we suspect. The most 

 interesting of these cases, from the standpoint of this study, is that 

 described by Blackman ('05) in Scoloyendra, a myriapod in which two 

 kinds of spermatocytes (and spermatids), differing remarkably in size, 

 occur. Two sizes of sperms are produced, and in general the case 

 reminds one strongly of 'polymegaly' in the pentatomid germ cells. 

 There is, however, one noteworthy difference in the arrangement of 

 the large and small generations, which in Scolopendra are not confined 

 to particular follicles or cysts, but occur side by side in all parts of the 

 testis. Blackman was inclined to regard the size differences as due to 

 differences in the supply of nutriment traceable to the spatial relations 



