STUDIES ON INSECT SPERMATOGENESIS. 71 



Whatever the origin of these vacuoles may be, they increase 

 rapidly in number, and soon become arranged in thread-like rows 

 running parallel to the future long axis of the nebenkern, subse- 

 quently becoming more or less fused together to form structures 

 reminding one of strings of beads (Fig. n). In cross-section, 

 these threads appear as single vacuoles, the position of which 

 marks the distribution of the threads (Fig. 12). In immediately 

 subsequent stages (Figs. 13 and 15) these beaded threads be- 

 come still more conspicuously developed, and as the chromophilic 

 substance disappears they become distributed throughout the 

 chromophobic envelope as shown by cross-sections of the neben- 

 kern (Figs. 2iA and B and 22). 



The account, thus far, is taken from Benda preparations, in 

 which the vacuoles stain faintly as described above. Occasion- 

 ally, however, the substance itself of the vacuoles stains more or 

 less darkly with the crystal violet (right half of Fig. n), becom- 

 ing much more conspicuous than is usually the case. Very fre- 

 quently these vacuoles are not demonstrated at all by the Benda 

 method, the chromophobic material appearing entirely clear. 

 The above account applies particularly to the large-cell genera- 

 tions; in the small-cell generations (in Murgantia the large-cell 

 generations also are not sufficiently large to make any difference 

 in this respect) the results with Benda are quite different. In 

 these the crystal violet often acts rather capriciously, but as de- 

 scribed by many authors it usually stains the zvholc mass of the 

 nebenkern an intense violet. In other words, the chromophobic 

 material takes the violet stain. This difference in staining be- 

 havior of the large and small nebenkerns, so fortunate for this 

 investigation, is probably due to some physical factor which re- 

 sults in the ready extraction of the stain from the large neben- 

 kern masses while from the small ones the stain is usually ex- 

 tracted with more difficulty. In the latter case, as in the former, 

 the substance of the newly formed vacuoles fails to take the violet 

 stain in good preparations, and thus the substance of the neben- 

 kern appears to be composed of two materials of very different 

 staining capacities. I thought, at first, that the material which 

 thus takes the crystal violet (in the small-cell generations) was 



