82 PROCEEDINGS: BOSTON SOCIETY NATURAL HISTORY. 
plate 16, figure 107. I feel obliged to consider this stage as the first 
maturation division for reasons which appear in the description of 
the second maturation division (see fig. 117). 
In this division the chromosomes never form such a perfect equa¬ 
torial plate as they do in corresponding phases in division of the 
spermatogonia (pi. 15, fig. 66). In nearly all cases some of the chro¬ 
mosomes are slightly eccentric. 
Metapliase .— The number of chromosomes in the equatorial plate 
is readily determined in these giant spermatocytes. I have repeatedly 
counted twenty-eight chromosomes. Sections through the equator 
of the spindles are common (pi. 16, fig. 110). The spindle is circu¬ 
lar in transverse section. In this first maturation division, there is 
no separation into mantle fibers and central spindle. The chromo¬ 
somes are regularly distributed, a fact which facilitates the counting 
of them. In many if not in all cases the chromosomes in the equato¬ 
rial plate show a marked tendency towards a lineal arrangement (pi. 
16, fig. 110), more marked in some cases than in others (pi. 15, fig. 
72). There is also a distinct radial arrangement of the granules of 
the cytoplasm, which represent the cut ends of the astral rays (pi. 
15, fig. 72; pi. 16, fig. 110). But this is not nearly so marked as in 
the case of the aster looked at from the pole (pi. 15, fig. 73; pi. 16, 
fig. 110). 
While the chromosomes still occupy the equatorial plate they split 
into halves. 
Anaphase .— The halves of each chromosome then move to oppo¬ 
site poles. The movement of some segments is faster than that of 
others, the result being that in this phase, also, it is a rare occurrence 
to find such perfect diasters as in a corresponding phase of sperma¬ 
togonia mitosis. 
They all reach their respective poles where they become crowded 
closely together into a more or less crescent-shaped mass partly encir¬ 
cling the centrosome (pi. 16, fig. 113; pi. 15, fig. 74). From what I 
have seen of this in hundreds of cases both in the living dividing sper¬ 
matocyte and in the best preserved material, I have been led to the 
conclusion that the chromosome segments approach the centrosome 
no nearer than the second microsome ring (pi. 16, fig. 109), which is 
the limit of that portion of the spindle which is formed from the 
nucleus. I feel certain that this is also the case in the karvokinesis of 
the spermatogonia (pi. 15, fig. 67). 
