142 A CYTOLOGICAL INVESTIGATION OF SOME SPECIES 
How the unpaired chromosomes behave during anaphase is not al- 
ways clear; probably most of them split longitudinally, the halves mo- 
ving towards the poles, each to one side. Occasionally however they 
seem to fall to pieces, while of course the possibility is not excluded 
that they reach once in a while, in an undivided condition, one of the 
poles. The important question was to show with certainty the presence 
of unpaired chromosomes and their presence in diakinesis, in meta- 
phase and in anaphase is now established. 
The study of nuclear plates of G. N. G., viewed from the poles, does 
not always allow us to distinguish between gemini and univalent chro- 

Fig. 40. 
Fig. 40. Anaphase of the heterotype division of G. G. N. G. (irregular division) 
x 2300. 
mosomes. The chromosomes are frequently irregularly arranged, often 
also the chromosomes of some of the gemini have already separated 
and show the first stage of anaphase, while univalent chromosomes 
may lie a little above or below the aequatorial plane. This makes coun- 
ting difficult, but a little persistance always reveals the presence of 
plates suitable for the determination of the chromosome number. 
Counting these, we get different numbers : 40 or more, but I have never 
met with less than 40 chromosomes. The number 40 itself also could 
but rarely be made out with certainty. Fig. 41 shows a nuclear plate 
with 39 chromosomes in the aequatorial plane, part of which show the 
first signs of anaphase, which however by the arrangement of the chro- 
