184 BOTANICAL GAZETTE [SEPTEMBEK 
Fig. 12 is another cell in telophase. One nucleus is uncut, show- 3 
ing 10 bivalent chromosomes. The other was sectioned by the knife. — 
Fig. 13 shows 11 chromosomes and a number of very small nucleoli. — 
Usually only one or two larger nucleoli are present. In fig. 14 there — 
are 12 bivalent chromosomes. In such cases g appear at the opposite 
end of the spindle. Figs. 17, 18 are very early telophases just after 
the nuclear membrane has been formed. The daughter nuclei grow 
very quickly to the size of figs. 19, 20, each of which shows 11 bivalent 
chromosomes, while in the nucleus represented in fig. 21 there are 
only 9. In general I have found that the nucleus having 9 chromo- 
somes is likely to be appreciably smaller than one having 11. Evi- © 
dently the amount of karyolymph secreted by a nucleus is a function 
of its number of chromosomes. This is also shown by the many cases 
of single chromosomes left behind in the cytoplasm and forming small 
nuclei during reduction, as observed in various forms, especially in 
hybrids. : 
These drawings of telophases are nearly all from nuclei which are 
uncut by the knife in sectioning. The sections are usually 10 # thick, 
so that in a majority of cases the nucleus is completely contained in 
one section. The nuclear membrane is conspicuous; hence PY 
focusing it can be determined with certainty that the membrane is 
uncut and that the contents of the nucleus are intact. All the nuclei 
in which counts have been made have been first shown to be intact 
in this manner. Moreover, whenever the chromosomes in both 
daughter nuclei could be counted, their sum was invariably found to 
be 21, whether 10+ 11 or, as occurred in a few cases, 9+ 12: 
There are a number of interesting features about the telophase, 
which I have studied with particular care. In the first place it is 10 
the best stage for counting the chromosomes with absolute accuracy, 
and is even better than diakinesis, because of the reduced number 
of chromosomes. In my last paper on reduction (11) I showed that 
pairing before the reduction division only takes place to 4 limited 
extent, and that the shapes characteristic of heterotypic chromosomes 
are therefore usually absent at this time, as will be seen from figs: 
26-34 of the paper referred to. During the period of interkines!> 
however, the entire chromosomes of the heterotypic mitosis haviS 
split to form bivalents in which the two parts are closely held together, 
Wie te. ‘i 
fa ia 
ee, eee eet 
