128 LOUISE B. WALLACE. 



equatorial plate prior to the second maturation division, their 

 V-shape becomes evident. The apex of the V is the point of 

 attachment of the spindle fibers, and the free, sinuous arms ex- 

 tend in various directions away from the spindle axis, giving a 

 bushy appearance to the mitotic figure (Fig. 32). 



On account of this arrangement of the ordinary chromosomes 

 it is now extremely difficult to identify the accessory chromo- 

 somes. Also the fact that they are present in only half of the 

 secondary spermatocytes lessens the chances of finding sections 

 cut in a favorable plane for their identification. In spite of these 

 disadvantages they can in some cases be clearly seen at the 

 equator of the spindle where they lie near together and at right 

 angles to the spindle axis. Sometimes indications of the longi- 

 tudinal split can be detected (Fig. 33). 



The arms of the V-shaped chromosomes shorten and thicken 

 while they also become straight and densely staining. The plane 

 of division passes through the apex of the V's and the rod-like 

 arms move to the opposite poles. The accessory chromosomes 

 divide along the line of the longitudinal split and their resultant 

 halves pass to the opposite poles a little more slowly than the 

 ordinary chromosomes. They are also distinguishable by their 

 larger size. It is now apparent that one half of the spermatids 

 will be " favored cells," containing two accessory chromosomes, 

 while the other half will not be favored (Figs. 34-39). In polar 

 views of the anaphase attempts were made to determine the 

 chromosomal number but after the utmost care I can give only 

 the probable number. In the majority of cases twenty-five 

 chromosomes were counted (Fig. 37), and their straight, rod-like 

 form makes it improbable that any of them were counted twice. 

 Occasionally twenty-four, twenty-six or twenty-seven were 

 counted, all of them appearing to be ordinary chromosomes. It 

 seems now as if the reduced number must be at least twenty-five, 

 instead of nineteen as given in my previous paper. In telophase,- 

 before the nuclear membrane forms, the ordinary chromosomes 

 again become slightly sinuous in outline. The daughter cells 

 must often move through a considerable arc in the process of 

 separating, as sections are found showing two cells not yet com- 

 pletely seperated and at the same time showing approximately 



