INSECT SPERMATOGENESIS 613 



fact that occasioiiallv (in C alios a mi a) cysts are found at 

 about the age of fig. oB (or later), in which each sperm head is 

 enclosed in a clear vacuole — presumably the vacuoles noted 

 above which have perhaps failed to be formed in a normal 

 manner. Something of this appearance is shown by Cook 

 (1910) in her figs. 132 and 134 of Automeris. This opinion 

 is further strengthened l)y the fact that in degenerate (apyrene) 

 sperms of Callosamia I have found that each head (now 

 moving back in the tail region) is accompanied by a droplet 

 of non-staining material (see Munson, 1906, fig. 49), over 

 which the acrosome passes. If these vacuoles are a product of 

 the nucleus we should expect just such a disposition of them 

 in the apyrene sperms ; but their connexion with the nucleus 

 is not easily accounted for on any other explanation of their 

 origin. In normal cysts, as the sperms grow older, these 

 vacuoles seem gradually to disappear, but their exact fate has 

 not been traced. 



The GolctI Apparatus and Acrosome. 



My chief interest in examining spermiogenesis in Lepidoptera 

 was centred on the origin and development of the acrosome, 

 especially in view of the account given by Gatenby (1917 a) of 

 the role of the Golgi bodies (his acroblasts) in this process. 

 According to Gatenby, all the Golgi bodies become swollen into 

 vesicular spheres during the early spermatid stages, and these 

 spheres fuse to form the basis of the acrosome. In each of 

 these spheres there is differentiated a small, darkly-staining 

 granule, which is also involved in the construction of the 

 acrosome itself. In a subsequent statement Gatenby and 

 Woodger (19'21) say, ' Our recent observations ... on several 

 other moths (e.g. Bis ton) have shown that in these insects 

 much of the apparatus finally passes as isolated crescents, 

 spheres, or dictyosomes into the elongating tails of the sperma- 

 tozoa.' The problem is thus left in a very unsettled condition. 

 In my previous papers on spermatogenesis I have endeavoured 

 to show that the acrosome is a product of the Golgi apparatus 

 plus idiosome, but that neither of the latter structures is made 



