INSECT SPERMATOGENESIS 597 



properly. The P.w.a. mixture gave the best general results 

 and was particularly valuable for studjang the Golgi bodies 

 and acrosome ; while for the mitochondria in spermatocyte and 

 early spermatid stages only Champy was satisfactory. Fe- 

 haematoxylin, sometimes with light green as a counter-stain, 

 was employed exclusively for staining. 



The Phenomenon of Polymegaly in Lepidoptera. 



In his first paper on Pygaera, Meves (1900) called atten- 

 tion to the fact that the spermatocytes are of two sizes, the 

 larger of which produces normal spermatids and sperms, 

 while the smaller undergoes abnormal maturation divisions 

 and produces abnormal (apyrene) sperms. The problem stated 

 in this simple form by Meves has been complicated by the 

 account of Munson (1906), who finds in Papilio two sizes 

 of spermatocytes and spermatids both of which develop in 

 a perfectly normal way. He Ijelieves that the small generation 

 is to be considered ' normal ' in size, since the large generation 

 only makes its appearance late in the life of the butterfly. 

 Munson unfortunately mixed up normal and abnormal stages 

 in his account of sperm formation, and his statements are 

 accordingly difficult to evaluate. Without at first recalling 

 either of these cases, I noted independently that the spermatids 

 in Callosamia are of at least two (possibly more) well- 

 marked sizes. Of these, the larger generation certainly gives 

 rise to normal sperms. On the other hand, the smaller sperma- 

 tids apparently give rise as a rule to abnormal sperms, but 

 nevertheless they are often found in an advanced stage of 

 normal sperm formation, and considered separately would 

 certainly not be thought abnormal. However, Gatenby (1917 &) 

 states that the abnormal condition which results in the forma- 

 tion of apyrene sperms may exert its influence at different 

 times, and it seems probable, therefore, that these small 

 spermatids in Callosamia would undergo degeneration at 

 a later stage ; and cysts of small spermatids in later stages 

 have in fact been found in course of changes possibly degenera- 

 tive in nature. This being the case, my results would coincide 



