40 H. L. WIEMAN. 



area of the second spermatocyte is rather distinctly marked off 

 from the rest of the cytoplasm (Fig. lo) and suggests that the 

 nuclear membrane disappears very slowly. 



Polar views of the metaphase display, as nearly as could be 

 determined, 12 chromosomes, presumably the haploid number. 

 It would seem that each chromosome is divided quantitatively 

 by a longitudinal splitting; although it must be remembered that 

 attempts at verifying this conclusion by studying the constit- 

 uents of the daughter groups are not satisfactory owing to the 

 tangled condition of the chromosomes. 



I find nothing resembling the small stained body which in 

 Neuroterus according to Doncaster passes undivided to one of 

 the spermatids. As Wilson has observed this body is of the same 

 nature as the chromatoid body seen in the growth-period and 

 s'permatocyte-division of Pentatoma. The chromatoid body is of 

 rounded form, dense arid homogeneous consistency, and after 

 double staining with haematoxylin or safranin and light green 

 is at every stage colored intensely blue-black or brilliant red, 

 precisely like the chromosomes of the division period or the chro- 

 mosome-nucleoli of the growth period. Nevertheless Wilson 

 finds that the body is neither a chromosome nor any kind of a 

 chromosome and takes no visible part in the formation of the 

 spermatozoa. In the transformation of the spermatids it wanders 

 far into the sperm-tail and is at last cast off altogether. 



I have not yet had opportunity to study the maturation phe- 

 nomena of the egg in either generation of Dryophanta, but obser- 

 vations confined to individuals of the bisexual generation point 

 to general conclusions which differ somewhat from Doncaster's 

 views regarding the chromosomal relations in the alternate 

 generations. In the material at my disposal spermatogonial 

 divisions are not abundant enough to determine the number of 

 chromsomes. While mitoses abound in the somatic cells of male 

 larvae and pupae, it is difficult to find good clear metaphases; but 

 wherever counts were possible, the number found was 12 (Fig. 

 19). In the follicle cells of the ovary I have found it less difficult 

 to count the chromosomes. Figs. 20, 21 and 22 are drawings of 

 metaphase plates of such cells in which the numbers are re- 

 spectively 13, 14 and 13. 



