58 
EDMUND B. WILSON 
3. An odd or accessory chromosome is present, and also a 
third m-chromosome. 
A study of the maturation process decisively establishes the 
third of these possibilities as the fact. 
h. The first spermatocyte-division 
Both testes contain immense numbers of both spermatocyte- 
divisions in all stages, and many of the cysts show the facts with 
great beauty. The first division itself at once indicates the true 
interpretation of the spermatogonial groups; and this is consis- 
ently borne out by the stages which precede and follow. 
In polar views (fig. 2, d-g) the first division metaphase is iden- 
tical in appearance with that of Anasa, Chelinidea, Narnia, and 
other coreids that have 21 spermatogonial chromosomes (including 
Montgomery's individuals of M. terminalis) . Eleven chromo- 
somes appear, including one very small central one surrounded 
by a ring of nine much larger ones, while the eleventh usually 
occupies a position outside the ring, as in figs. 2d, 2/, (figs. 2e, 2g 
are given to show exceptions to this). From these views alone 
we should infer that the spermatogonial number is 21, that the 
small central chromosome is the m-bivalent, and that the eccentric 
one is the accessory. This will appear upon comparison with 
figs. 2, h, i, which show two corresponding views of Montgomery's 
material of M. terminalis. Side views at once reveal the fact, 
however, that the central body in No. 64 is not a bivalent but a 
triad element, consisting of three small chromosomes united end 
to end (figs. 2h, 3a, h,) and it is perfectly evident that these are 
identical with the three very small ones of the spermatogonial 
groups. Hundreds of these figures have been observed, in almost 
all of which the three components have the linear arrangement 
just described; but now and then a different grouping occurs, as 
may be seen in both side (fig. 3c) and polar views (fig. 2g). 
In the ensuing division the ten larger chromosomes divide 
equally, showing as they draw apart the curious forms represented 
in fig. 4, which are closely similar to those described in Anasa 
by Paulmier ('99). As the figures show, the chromosomes in 
