Studies on the Cytology of some Species of Taraxacum. 
457 
like or spindle-shaped body, which is connected with the thin portions 
of the thread, thns presenting more or less an appearance of pearl-string. 
Here the chromatic thread is without doubt single, and this fact is in 
sharp contrast to that seen in T. platycarpum, where the double nature 
of the thread is very clear. (Compare Figs. 47 — 48 with Figs. 7 — 8.) 
The condensation and contraction, which has been going on in the 
spirem, continnes tili now, and the delicate fibres which connect the 
Segments are gradnally drawn apart and finally disappear. This latter 
action does not take place simultaneously throughout the whole length 
of the thread, but here and there in various times, tlius forming gronps 
of the chromosomes (Fig. 48). Each group is composed of a few segments 
*arranged end to end. Finally the thread splits into a definite number 
of chromosomes, which are short rod-like or spindle-shaped bodies (Fig. 49). 
The condensation and contraction proceed in these chromosomes, nntil 
they become more compact, granulär or spherical bodies. Fig. 50 shows 
a pollen-mother-cell cut in two serial sections and here one chromosome 
and one nucleolus were brought a little way out of the nucleus bv the 
knife edge. The nucleus is now in diakinesis stage, each chromosome 
being scattered at the periphery of the nucleus. In many cases the pairing 
of chromosomes is very obscure and they are distributed nearly at equal 
distances from one another, but in other cases the double structure of 
the chromosome is somewhat recognizable, though it is not so distinct 
as in the case of T. platycarpum. After a short time there appear the 
fibrillae in cytoplasm around the nucleus and they enter into the nuclear 
cavity with the breaking down of the nuclear membrane, thus producing 
a multipolar spindle (Fig. 51); the large nucleolus also disappears during 
this process. The multipolar condition does not dure for a long time, 
but soon the spindle becomes changed into a typical bipolar one, as in 
Fig. 53. 
In the first mitosis just mentioned, there are two cases. In the 
first case, the chromosomes arrange themselves very regularly on the 
equatorial plate of the spindle, which appears very wide (Figs. 52, 53). 
At this stage or earlier I made attempts to count the number of chromo- 
somes in few sections, but the exact determination was very difficult 
on account of their great number and close aggregation (Fig. 52). The 
number of chromosomes, which was counted in the polar view of the 
equatorial plate or in diakinesis stage, appears, however, to lie between 
36 — 40. At the anaphase of the first mitosis the daughter-chromosome 
at both poles of the spindle was counted and it proved that there exists 
also nearly the same number of them, and moreover this number coin- 
