274 ELIZABETH A. SMITH. 



the difference in appearance. The cytoplasm in some cells 

 was clear and homogeneous, while in others it was alveolar 

 and in most cases formed only a narrow sheath around the 

 nucleus. Fig. 4 shows groups of spermatogonia with the struc- 

 tures just described, nucleolus, chromatin body, nuclear net- 

 work and alveolar cytoplasm. Fig. 5 represents a spermato- 

 gonial cell from a smear preparation in which the chromatin 

 granules appear as clumps near the nuclear membrane. Fig. 6 

 shows the 25 spermatogonial chromosomes in an aceto-carmine 

 smear. 



In the spermatogonial divisions, the undivided chromosomes 

 could rarely be distinguished, as all the chromosomes in the 

 metaphase usually blend into a black compact mass. In Fig. 7, 

 a polar view of the metaphase before the last spermatogonial 

 division, a few of the separate chromosomes can be seen. They 

 are dumbbell-shaped, varying in size and are not as large as 

 those found by McGill ('04), in Anax junius in the same stage. 

 In some of my own preparations of Anax junius there were 

 clear polar views in which 27 chromosomes could be counted. 

 This verifies the corrected count of Lefevre and McGill ('08). 

 In Sympetrum semicinctum., however, it was impossible to find 

 many cells in which the chromosomes were sufficiently separated 

 to permit of a satisfactory count. In many instances where 

 partial counts were possible, more than twenty could be dis- 

 tinguished. Fig. 8 shows twenty-five plainly in a polar view 

 of a telophase, and while the two cells in which twenty-five could 

 be distinctly counted do not afford sufficient evidence from 

 which to draw a conclusion, judging from the reduced number 

 found in the primary and secondary divisions, the correct 

 spermatogonial number should be twenty-five. The dense 

 appearance of the polar view is due to the deeply staining pro- 

 clivity of the threads which connect the chromosomes. 



Figs. 9, 10, and n are telophase stages of the last spermato- 

 gonial division, showing approximately the amount of chromatin 

 going to each cell. The telophase stages are more abundant than 

 the metaphase ones in my material. A polar view of the telo- 

 phase shows 25 chromosomes distinctly in one case and many 

 show plainly 21 chromosomes and there are indications of others 



