SPERMATOGENESIS OF STENOBOTHKUS VIRIDULUS. 



explain the resolution of the chromosomes into spiremes or reticulum, and 

 their later shortening and consequent closer association of granules — the 

 chromatin in this case serving merely as a conA^enient vehicle for the precise 

 distribution of character factors, or their equivalents, between the two 

 daughter cells. 



Material and Methods. 



My material was collected at Nannerch, in Flintshire, N. Wales, in the 

 last week of August 1909. The grasshoppers were killed in chloroform 

 within a few hours of capture, and were placed whole in the fixative after the 

 wings and legs had been removed, and the integument of the back slit up to 

 allow readier access to the fluid. I have obtained excellent results with 

 Perenyi's chromo-nitric acid solution, the resting-stages and various phases 

 of mitosis being very perfectly preserved : the majority of writers on insect 

 spermatogenesis, however, appear to have used Flemmiug's strong chromo- 

 aceto-osmic acid solution, Hermann's platino-aceto-osmic acid solution, or the 

 fixatives of Bouin and Zenker. 



The grasshoppers were transferred after two hours to a 50 % aqueous 

 solution of alcohol, and an hour later were placed in a 70 % solution, in 

 which they remained for twelve hours ; they were then stored in a solution 

 of 80 % alcohol. This storage solution was changed twice during the first 

 month, having become thick and discoloured with pigment. 



When required for embedding, the testes were dissected out, and placed 

 for twenty-four hours in a 90% solution of alcohol: after being passed 

 through absolute alcohol and cleared in cedar-wood oil, they were embedded 

 in paraflin, remaining for twenty minutes in the first bath and for fifteen in 

 the second. I used paraffin with a melting-point of 52° C, since I found 

 that paraffin with a higher melting-point had a tendency to overheat the cells. 

 Sections were cut with an ordinary Cambridge rocking microtome to thick- 

 nesses varying from 5 to 10 [m, and were invariably stained on the slide. 

 The nuclear stains used were Heidenhain's iron hsematoxylin, iron brazilin, 

 and safranin, the first-named being used alone or in conjunction with a 

 plasma stain — e. g., eosin, congo-red, or picro-carmine ; I also used the 

 tricolor stain of Flemming, and the permanganate of potassium method of 

 !FIenneguy. 



In staining with the iron hsematoxylin, 1 used, as a mordant, an aqueous 

 solution of iron alum, in which the slides remained for six hours ; they were 

 then stained for twelve or fifteen. Davis left his slides in the mordant for 

 only two hours, and in the stain for from four to six ; but 1 have found that 

 the longer period gives better results as regards sharp definition, while the 

 process of differentiation can be more perfectly controlled. In the cases 

 where a second stain was used, the slides were left for ten minutes in the 

 plasma stain before being transferred to the iron hiematoxylin : the iron 



