Davis. — Cytological Studies on Oenothera . /. 557 
become the spirem. This process may be readily followed and is illustrated 
by Figs. 12-20, which show that the threads of the older stages are 
more than twice as thick as those of the younger, and that the total 
length of the threads (so far as they may be traced) becomes very much 
less as synapsis proceeds. 
The forms of the fully contracted synaptic knot are exceedingly 
varied. In the most extreme cases the threads are drawn tightly together 
into an irregular lumpy mass (Figs. 21 and 23). More often, however, 
there are several well-defined loops extending into the nuclear cavity from 
the dense contracted centre of the synaptic knot, as illustrated in Figs. 19, 
20, and 23. The number of loops is not fixed, and they are distributed 
irregularly and are of various lengths. The loops that extend to the 
periphery of the nucleus are attached to the nuclear membrane, which 
is frequently drawn inward at such points probably because of the pull 
exerted by the contracted threads. Even when the contracted material 
is in the form of a dense mass (Figs. 21 and 22) it is always possible 
to distinguish definite lobes that in earlier stages were undoubtedly loops, 
but are now tightly drawn against the synaptic knot. 
The mid-phase of synapsis, or the period of greatest contraction, is 
illustrated by Figs. 21, 22, and 23, which are fairly typical of various 
conditions. It should be noted that the nuclear cavity contains only the 
chromatic mass, the single large nucleolus (rarely two large nucleoli), 
occasionally smaller nucleoli (Fig. 21), and perhaps remnants of the ex- 
tremely delicate threads (Fig. 23) which are present close to the periphery 
in earlier stages of synapsis (Figs. 13, 15, and 19). The chromatic mass 
may lie near the centre of the nucleus (Fig. 22), but it is usually some- 
what pressed against the large flattened nucleolus and consequently at 
one side. The latter situation suggests a polar organization of the 
nucleus that is frequently emphasized by the radiating arrangement of 
the loops from the chromatic mass as a centre (Figs. 20 and 23). The 
chromatic loops are very important structures, for some of the ring-shaped 
bivalent chromosomes arise directly from them. 
The Formation of the Bivalent Chromosomes. 
As stated before, the structure of the synaptic knot when fully con- 
tracted is very difficult of analysis. The chromatin frequently has the 
appearance of a lumpy mass (Figs. 21 and 22), the surface being broken 
by irregular globular projections, some of which were formerly the loops 
that extended into the nuclear cavity. In other cases these loops may 
remain clearly defined, although much shorter than in earlier stages of 
synapsis (Fig. 23). Differential staining clearly indicates that the con- 
tracted synaptic chromatin has regions of greater density, some of which 
