STUDIES ON INSECT SPERMATOGENESIS. 75 



aggregates, at the same time absorbing the chromophobic material, 

 with the final result that the axial filament lies naked in the cyto- 

 plasm with the fat droplets scattered along its course. These 

 later pass out along the tail to be collected in the protoplasmic 

 masses that are sloughed off in the final maturing of the sperms, 

 which, as a result, possess no trace of mitochondria or their de- 

 rivatives. This again reminds one of Holmgren's account, and 

 the same comment relative to the findings of other authors is ap- 

 plicable by way of criticism. 



In view of the unusual fate ascribed by Holmgren and Vejdov- 

 sky to these final structural components of the nebenkern, it is of 

 interest, finally, to inquire into their later history in the Hemip- 

 tera. In stages subsequent to that of Fig. 26, the Cajal method 

 fails to impregnate the central substance of the mitochondrial 

 sheaths, which continue to draw out rapidly with a consequent 

 decrease in their diameter. The only evidence which I have been 

 able to get on the structure of the nebenkern in these later stages 

 has been obtained from Benda preparations. I have already 

 called attention to the fact that in some cases (e.g., Murgantia) 

 I have obtained preparations in which the outer, chromophobic 

 material- stained intensely with crystal violet, while the substance 

 of the vacuoles (the central substance) remained colorless. Ac- 

 cordingly, when the substance of the latter has run together to 

 form a single axial core for each half of the nebenkern (Fig. 26), 

 the two substances can still be distinguished, the outer layer (the 

 chromophobic portion) staining a deep purple while the central 

 core (the central substance) is unstained. (See Bowen ('22) 

 Figs. 76 and /8.) 1 As I have described in another paper (Bowen, 

 ('22)) the mitochondrial sheaths begin now to develop character- 



i It may be as well at this point again to call the attention of the reader to 

 the confusing differences in the staining behavior of the nebenkern substances. 

 In the first place, it is to be noted that the central substance almost invariably 

 fails to take the crystal violet of Benda's method, while the chromophilic 

 material is always stained violet if the nebenkern retains the stain at all. The 

 chromophobic material, however, behaves in a variable way. Under the best 

 conditions it does not retain the crystal violet (at least in early spermatid 

 stages) when the rest of the preparation is properly differentiated ; but some- 

 times in the early spermatid stages (small-cell generations), and invariably in 

 the later ones, it stains heavily with the violet an exact reversal of staining 

 behavior. 



