178 PROCEEDINGS OF THE ACADEMY OF [Feb., 



an X is then obviously formed from a V, not by any extension of the 

 longitudinal split, but simply by the point of contact of the two uni- 

 valent chromosomes shifting its position. The decisive phenomenon 

 through these prophases is that most of the chromosomes preserve their 

 original forms of bent rods, or modify them into straight rods, leading 

 toward the forms most frequent at the end of the prophase (fig. 52), 

 where the earlier point of union of two univalent chromosomes is 

 recognizable (x), and the longitudinal split sometimes still discernible. 

 For the greater number of the chromosomes the changes of the pro- 

 phases lead to the retention of approximately their original form, but 

 with a gradual partial or complete closure of the longitudinal split; 

 and there is no reason to hold that the longitudinal split ever widens 

 and remains open in such a manner as to change the position of the 

 long axis of a bivalent chromosome. The heterochromosome under- 

 goes no marked modification during the prophases; at first each of its 

 univalent portions shows still the longitudinal split, shown in end view 

 on fig. 49 (A^. 2), but toward the close of this period this split appears 

 to close up. 



In the equatorial plate of the first spermatocyte (figs. 53, 54) are 

 found thirteen larger bivalent chromosomes, and sometimes a minute 

 chromatin body (S.) which does not appear to be bivalent, at least it 

 is not bipartite. The latter may represent one of the tw^o minute 

 chromosomes of the spermatogonia (fig. 41) ; one of these small bodies 

 is occasionally found in the monaster stage of the second spermatocytes 

 (figs. 62, 63). Their behavior in the growth period could not be deter- 

 mined, so we must disregard them in our analysis of the chromosomal 

 relations. The first maturation figure has then thirteen bivalent 

 chromosomes, corresponding to the twenty-six larger univalent ones 

 of the spermatogonia. Lateral views of the first maturation spindle 

 are shown in figs. 55-5<S. "Wliile on most of the chromosomes at this 

 stage the longitudinal split is not evident, in some cases it is still per- 

 sistent, as notably in those lettered K in figs. 55, 56, 58. Particularly 

 the one in the last figin-e is valuable in demonstrating how the chromo- 

 somes become placed in the spindle: the point of junction (x) of the 

 two univalent components lies in the equator, therefore the one uni- 

 valent chromosome just above and the other just below this plane. 

 From this arrangement and from the mode of insertion of the mantle 

 fibres it is evident that in this mitosis the two univalent chromosomes 

 of each pair become separated from each other into opposite cells, and 

 that this is a reduction division. Generally the long axis of each biva- 

 lent chromosome is parallel to the axis of the spindle, which is always 



