in others by a sharply
acute angle. In a slightly later stage (fig. 43), when all of the short
loops have straightened and united in pairs, the point of union is no
longer visible, all of the loops being rounded at the bend and of equal
thickness throughout. My attention was first called to this method of
synapsis by the conspicuous difference in number and length of loops in
the synizesis stage compared with the later bouquet stage just before
the spireme is formed. Following the synapsis stage shown in figure 43
comes one in which the loops lose their polarized arrangement and unite
to form a continuous spireme (figs. 44 and 45). In this form, the
heterochromosome pair could not be distinguished until the spireme
stage, and it is, therefore, uncertain whether these chromosomes remain
condensed after the last spermatogonial divisions and are hidden among
the massed and deeply staining loops of the synizesis and synapsis
stages, or whether they pass through the same synaptic phases as the
other chromosomes, condensing and remaining isolated at the beginning of
the spireme stage. An early prophase of the first maturation mitosis
(fig. 46) shows segments of the spireme longitudinally split, and in
some cases transformed into crosses which show a transverse division
also. Most of the equal bivalents have the dumb-bell form in the
spindle (figs. 47-49). One is ring-shaped, the ring being formed by
union of the free ends of the segment so that the spindle fibers are
attached to the middle of each univalent chromosome (fig. 49). This
method of ring formation, like that described by Montgomery ('03) for
the Amphibia, is of very frequent occurrence in the spermatocytes of the
Coleoptera. The dumb-bells are so bent at the ends (fig. 52) that the
spindle fibers, here also, are attached at or near the center of each
univalent component of a bivalent chromosome, and the separated,
univalent chromosomes go to the poles of the spindle in the form of Vs.
As in _Tenebrio_ the heterochromosome pair is late about coming into the
equatorial plate (figs. 47-48), but it does finally take its position
with the others (fig. 49) and separates into its component parts
somewhat earlier than the other bivalents (figs. 52, 53). Figures 50 and
51 show polar views of the metaphase, the smaller element (_x_) being
the unequal pair. The chromosomes in late anaphase are too much crowded
to give clear drawings. As in all the beetles so far studied there i