102 
THIRTEENTH REPORT. 
The tetrads show a variety of forms which may he grouped as rods, 
crosses and rings. Some of them are straight double rods with a slight 
opening in the center; others are so bent upon themselves as to appear 
y shaped; and still others are so curved as to resemble a twisted dough- 
nut. The most common form assumed is that of a ring. Sections well 
destained show very clearly the quadripartite structure of the rings. 
Figures 10 and 11 show some of the characteristic form of the tetrads. 
During the formation of the tetrads there can be distinguished in the 
nucleus a deeply staining, compact body, the accessory chromosome (Fig. 
10). In preparations which have been destained for some time the 
ordinary chromosomes present a rough, granular appearance but the 
accessory remains compact. 
At the time when the tetrads take on a homogeneous appearance, the 
nuclear wall breaks down and the chromosomes pass into the late pro- 
phase of the first spermatocyte division (Fig. 15). The primary sper- 
matocyte division is relatively short as compared with the growth period, 
but, on the other hand, it is relatively long when compared with the 
secondary spermatocyte division. At least one is lead to conclude this 
from the fact that a large part of the follicle is composed of cells which 
are in the growth period. Most of the actively dividing cells are primary 
spermatocytes and it frequently happens that considerable search is neces- 
sary in order to find secondary spermatocytes. The secondary sperma- 
tocytes can be distinguished by the characteristics of the mitotic figure. 
The chromosomes are smaller and appear in metaphase as dumb-bell 
forms on the spindle (Fig. 1G). 
When the autosomes of the primary spermatocytes take up their posi- 
tion in the equatorial plate the accessory chromosome moves toward 
one pole of the spindle. It may occupy various positions, either directly 
at the pole, at some place between the pole and the equator, or entirely 
outside the spindle as described by Baumgartner (1904) for Gryllus. 
Some of the various positions it takes up are shown in Figs. 12, 13, 14. 
The accessory is bean-shaped and shows a more or less distinct longi- 
tudinal splitting (Fig. 12). The concave side may be toward or away 
from the spindle. It does not divide in the first spermatocyte division. 
Occasionally this chromosome lags behind the others and appears as a 
large distinct element on the spindle after the other chromosomes have 
reached the poles (Fig. 13). It is a striking fact that cells which exhibit 
this peculiarity occur in groups, and are not scattered among the other 
types of cells. Up to the present time a large enough number of these 
cells have not been found to warrant separating them as a distinct type. 
In the metaphase the autosomes take up a position along the equatorial 
plate so that the long axis is at right angles to the spindle axis. As 
the mantle fibres contract the chromosomes are pulled apart and pass 
to the poles as loops (Fig. 14). The middle of each segment separates 
from its fellow first but the ends remain in contact for some time, the 
pull of the mantle fibres stretching out the two chromosomes in a plane 
at right angles to the original axis, thus giving, at first glance, the ap- 
pearance of a transverse division at this stage. However, careful trac- 
ing of the chromosomes from the tetrads to the anaphase indicates that 
the first division is longitudinal. 
