AND SPECIES HYBRIDS WITHIN THE GENUS SACCHARUM 125 
of this body; they can also be entirely covered by the nucleolus, thus 
becoming invisible. In fig. 12 this is probably not the case. The pictu- 
red nucleus shows 56 gemini: The case just described however is an 
exception and is to be considered as a transitional stage of very short 
duration which usually proceeds less regularly. The end-stage of diaki- 
nesis 1s frequently met with in our preparations, e. g. the stage in which 
the chromosomes which have formed the gemini are closely adpressed 
to one another (fig. 13). Many gemini show by their cross-shape plainly 
their double nature, but in the case of many others, especially in that of 
those which one observes from their narrow side, and which occur in the 
picture at the side of the nucleus, the bivalent nature is not clearly dis- 
cernable. The same situation is shown by the two chromosomes which 
are adpressed to the lower edge of the nucleolus and simulate excres- 
cences of this body. Here also one sees, that nearly all chromosomes 
are adpressed to the unclear wall, but that a few chromosomes, near 
the nucleolus, make an exception. The latter occurs frequently. In this 
nucleus the number of chromosomes again was 56, while it was pro- 
bable that all were bivalent, though this condition could not be shown 
in all of them. 
After diakinesis the nuclear wall disappears, the chromosomes arran- 
ge themselves in a heap, from which kinoplasmatic threads radiate 
into different directions, which together form a multipolar spindle. 
This stage was rather frequently observed, but never occurs in multi- 
tude simultaneously, which shows that it passes quickly. From the 
multipolar spindle a bipolar one arises, the gemini arrange themselves 
in the aequatorial plane. So the metaphase of the heterotype-division 
is reached. Observed from the side, the spindle threads can clearly be 
seen, but their connection with the gemini can but rarely be observed. 
The metaphase of Saccharum spontaneum is especially well fitted to 
count the chromosomes. It occurs frequently in the preparations and 
frequently simultaneously also, which shows that this stage lasts rather 
a long time *). In nuclear plates, viewed from one of the poles, nearly 
all gemini can be seen simultaneously, they are then round or angular 
1) Fig. 11 shows a crosssection of a part of an anther of Glagah K epandjen; here 
one sees 6 aequatorial plates simultaneously. To investigate metaphases cross- 
sections are best, because more nuclear plates occur perpendicular than parallel 
to the longitudinal axis of the anther. To study the diakinesis-stage longitudinal 
sections however are preferable. 
