320 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY. 



this is not the case. Furthermore, there are certain gradations between 

 the two. One such is represented in Figure 62, in which two of the 

 chromosomes have separated before the others, the latter showing the 

 elongate dumb-bell-like form. 



In seeking the conditions which determine the occurrence of one or 

 other of these two forms of spindles, we must, I think, turn to external 

 causes. The rarer of the two is found on sections near the base of the 

 gonad only, in the neighborhood of the mesogloea, occurring in isolated 

 cells surrounded by spermatogonia; while the more typical spindles 

 occur in groups in a more peripheral position, where primary and sec- 

 ondary spermatocytes, spermatids, and spermatozoa form a densely 

 crowded layer. From this fact it seems not improbable that the ap- 

 parent difference may be due either to the relative amount of pressure 

 to which the cells are subjected, or to the readiness of their access to a 

 food supply. It is interesting in this connection that similarly Down- 

 ing (:05) found in Hydra the occurrence of two kinds of mitosis in 

 the spermatogonia, a deviation from the common type being seen in cells 

 which divide near the margin of the spermary, where the process more 

 nearly resembles that of interstitial and ordinary ectoderm cells. The 

 final result, however, was the same in both. 



Anaphase. — As the chromosomes approach the poles they become, as 

 in the last spermatogonial division, closely compacted together (Fig. 68) ; 

 yet careful focusing always resolves the mass into separate component 

 bodies. In the migration one chromosome often reaches the pole in 

 advance of all the others (Fig. 62) ; this, however, is not invariably 

 the case. 



The achromatic figure of the first maturation division is of the same 

 simple type as that seen in spermatogonia and in entoderm cells, its 

 most interesting feature being, as there, the total absence of astral radia- 

 tions. The spindle fibres are exceedingly delicate, and after the ana- 

 phase are no longer distinguishable as such, while the interzonal filaments 

 are stout, stain deeply, and in the telophase bear a series of prominent 

 Zwischenkorpor, each of which is clearly a thickening of one of the 

 filaments. There appears to be a single filament attached to each chro- 

 mosome ; consequently there are only one-half as many as in the preced- 

 ing cell generation. The centrosome, a single minute granule without 

 surrounding sphere, is first visible in the early metaphase, and cannot 

 be traced with certainty after the breaking down of the spindle fibres 

 in the anaphase. 



Before proceeding to describe the second generation of spermatocytes, 



