26 Overton.— On the Organization of the Nuclei in the 
chromosomes are usually straight, with the two parts lying very close 
together. Sometimes, however, they may be somewhat bent, but, as 
Fig. 14, PL I, shows, they are eventually straightened out by contraction. 
The substance of each univalent chromosome contracts upon its own linin 
thread to form shorter bodies, leaving a more or less distinct double linin 
thread behind. Such appearances as is shown in P'ig. 16, PI. I, proves that 
each univalent chromosome condenses in its own linin thread. (Compare 
also Figs. 18, 19, and 20, PI. I.) At this stage (Figs. 16, 17, and 18, PI. I) 
univalent chromosomes are strikingly like, in size, shape, and staining 
reactions, the prochromosomes which appear in earlier prophases. (Compare 
Figs. 3 and 5 and Pigs. 14 and 18, PL I.) The contraction of the chromo- 
somes to the form of parallel double rods, which they have on the 
equatorial plate, is very gradual. The bivalent chromosomes are thus 
formed by a contraction of the pairs of univalent chromosomes upon the 
linin threads, which are probably incorporated into the chromosomes 
together with the chromatin. It is thus clear that the bivalent chromo- 
somes are not formed by a bending or folding of the spirem as maintained 
by Farmer and Moore (’ 05 ). 
The bivalent heterogeneous spirem, which is shown in Figs. 14 and 15, 
PL I, segments, forming the bivalent diakinetic chromosomes. Since the 
two univalent portions of the bivalent spirem may often be separated from 
each other, the two portions of each bivalent chromosome are also often 
entirely separated after the spirem segments. The chromatic portion of 
each univalent part of these structures contracts somewhat, the bivalent 
chromosomes appearing as short, rather thick double rods (Figs. 19 and 20, 
PL I). Fig. 21, PL I, shows two bivalent chromosomes still bound together 
by the bivalent linin intervals. The segmentation of the spirem has 
apparently been delayed at this point, leaving the two bivalent chromo- 
somes still held together by linin. The four univalent chromosomes are 
distinctly visible. Such figures, which are not uncommon, show that the 
line of approximation (so-called longitudinal split) is the same as is found 
in the thick spirem. It is also evident from such figures that the stage 
represented in Figs. 22 a , PL I, could not possibly arise by any process of 
bending. The line which separates the two univalent chromosomes is 
therefore exactly the same line which has been visible since the two 
prochromosomes became associated in pairs in the early prophases. 
Beside the parallel double rods (Fig. 22#, PL I), which are the most 
usual figures found in Thalictrum purpurascens , other forms of figures 
occasionally occur. Fig. 22 #, Pl. I, represents an 8-shaped figure. 
Figs. 22 c , 22 d, and 22 e, PL I, are chromosomes taken from a late diakinetic 
stage. In Fig. 22 c, PI. I, the two univalent portions are so orientated 
as to form an O-shaped figure. In the portion of the O at the right of 
this figure a longitudinal split is indistinctly visible. Just before the 
