1922] SEARS—TARAXACUM 315 
Instead of occurring uniformly, thickening of the thread seems 
to be accompanied by the beginnings of its segmentation. Figs. 
g and 1o show nodes whose structure is seemingly homogeneous. 
On the other hand, it is always possible to find some showing 
clearly a longitudinal duality, even culminating in a divergence or 
forking of the two attenuate ends. If this appearance of doubleness 
were visible in every node and the divergent internodes were not 
visible, one would be justified in seriously questioning the validity 
of the interpretation. It might then be simply the lateral shadow- 
ing normal to translucent cylindrical bodies. If, however, there 
is really duality, the separating plane in certain nodes must lie 
more or less parallel to the section and hence not be visible. This 
would account for those nodes whose appearance is homogeneous. 
The nodes rapidly shorten and become truly homogeneous, only 
the bifurcate internodes remaining as evidence of the double origin 
of each chromosome. As in all chromosomes, there are occasional 
lateral projections in addition to the forking internodes, due 
doubtless to imperfect retraction of pseudopodia at some time 
during aggregation. The papers already cited (12, 17, 26) evince 
little proof of close attention to this stage, a circumstance doubtless 
due to its transient character. Figs. 11-13) show it in varying 
aspects. Fig. 12 suggests a rough correspondence between this 
phase and the so-called second contraction. Certainly the thread 
shortens greatly, and the chromosomes as they first cut apart are 
no longer peripheral, but in the nuclear interior. 
The bifurcations at each end of the chromosome are not retracted 
at once, but may shift slightly in position. This gives the appear- 
ance of pseudopodia, generally four in number. Stork, Osawa, 
and JuEL have all more or less plainly figured but not accounted 
for these pseudopodia. Fig. 14), as those of the authors cited, 
shows that after the chromosomes drift to the nuclear membrane 
and become peripherally oriented, the quadruple projections tend 
to move to the side of the chromosome away from the membrane. 
These, with other irregular projections of earlier or later origin, may 
constitute the “fringe” referred to by Juret and figured by Osawa. 
Counts at this stage, and also at the somewhat later one 
resembling diakinesis, show twenty-six chromosomes. This agrees 
