traced
beyond the equatorial plate of the latter mitosis.

(6) The similarity of all the normal spermatids, though one-half of them
must contain the element _x_, the other half not.

(7) The varying and often large number of degenerate spermatozoa.

An attempt was made to determine the somatic number of chromosomes. The
dividing cells of the follicles of young eggs seemed to afford the most
favorable material, but even here there was so much overlapping of the
ends of the chromosomes that it was impossible to be absolutely certain
of the number. In the two most favorable cases 23 were counted (fig.
94). This differs from McClung's count for similar cases among the
Orthoptera, and Sutton's for _Brachystola magna_. The eggs have so far
resisted all efforts to learn what part the odd chromosome may play in
fertilization.

Tenebrio molitor.

In the metaphase of all spermatogonial mitoses where it was possible to
count accurately, 20 chromosomes were found, 19 large ones of
approximately equal size, and 1 small spherical one (figs. 169, 170).
There is nothing in the resting nucleus of the spermatogonia which would
suggest either a nucleolus or an accessory chromosome. The chromatin
stains well during the whole growth period of the spermatocytes, but it
is impossible to separate the period into so definite stages as in most
other forms.

In the youngest spermatocytes one finds occasionally a cyst containing
cells with nuclei like those of figures 171 and 172, indicating that a
brief "synapsis" or condensation stage occurs at the close of the last
spermatogonial mitosis. During the greater part of the period the
chromatin forms a heavy, irregular, and often segmented spireme (figs.
173, 174). Shortly before the first maturation division, such split
segments as appear in figure 175 are sometimes found; some of these
simulate tetrads with slender connecting bands between the paired
elements. Again, one finds a few cases like figure 176, where the
spireme is segmented into bivalent chromosomes, each component showing a
longitudinal split. This figure also shows the small chromosome.
Usually, however, the irregular and much tangled spireme (figs. 173,
174) condenses into a heavy segmented band variously disposed in the
nucleus (fig. 177). This band soon separates into the bivalent
chromosomes shown in figures 178 and 179, giving 9 symmetrical pairs and
1 unsymmetrical one (fig. 179 _s_) composed of the small chromosome an