44 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY. 



however, in the anaphase, the fibres connecting the chromosomes and 

 centrosomes do shorten, and the result is that the chromosomes are 

 drawn toward the poles, and at the same time become apposed to each 

 other and form a compact mass. The centrosomes during all the changes 

 after their division in the anaphase of the first spermatocyte remain dis- 

 tinct and at no time show any signs of another division. They pass 

 over into the spermatid in this simple condition as small deeply stain- 

 able spherical granules. 



I have said that during the later stages of mitosis the cells elongate 

 verv much. Tins is so marked that in the telophase the length of 

 the cell is often as much as four times the average width. As elonga- 

 tion continues, a constriction of the cytotheca appears in the region of 

 the equator of the cell. This proceeds very much as in the first sper- 

 matocyte, and results in two cells, the dividing walls of which are inde- 

 pendent. These cells are united, as in the preceding division, by a 

 bundle of persisting spindle remnants, but here the structure is very 

 poorly developed and persists a very short time, being often destroyed 

 by the rotation of the daughter cells. 



By this division there arises from each second spermatocyte two 

 spermatids. These are rather large cells of an irregular oblong shape, 

 containing a relatively small nucleus. This nucleus is at first sur- 

 rounded by a very delicate membrane ; later, however, as the cell proceeds 

 in its growth, this becomes more marked. The chromatin at first is in 

 the form of definite bodies (the chromosomes), but later these become 

 granular without definite outline. When this is accomplished it is 

 seen that in one half the cells one of the chromosomes does not lose 

 its homogeneous character, but continues to stain as in the metaphase. 

 This is the accessory chromosome, which up to the anaphase of the second 

 spermatocyte was peculiar by reason of its dyad characteristics, while 

 the other chromosomes were plainly tetrads. During the first matura- 

 tion division this modified element has passed over undivided into one of 

 the daughter cells, while in the second division its component chromatids 

 have been separated. It is quite evident that by this unequal distribu- 

 tion of the chromatin two distinct kinds of spermatozoa necessarily arise. 

 This phenomenon has recently been reported by several investigators of 

 arthropod spermatogenesis, and several theories have been founded upon 

 this fact. I will not enter upon these here, but later, in the discussion 

 of the literature, will give some account of them. 



