510 THE CYTOLOGY OF THE SUGARCANE 
matine-elements should be present in the nuclear plates of the homoty- 
pe division, while in fact there is always a distinctly larger number. 
The halves of the univalent chromosomes consequently occur in a un- 
paired condition in these nuclear plates. ; 
Views of nuclear plates from the side, showed that in the homotype- 
division also, the chromosomes were far from accurately lying in the | 
aequator. Many chromosomes were seen outside of the aequator on the 
spindles. The anaphase also showed the irregularity of the homotype 
division clearly. In many cases chromosomes lagged behind on the ae- 
quator. In one anaphase, viewed from one of the poles, approximately 
80 chromosomes could be counted, these were partly situated in the 
groups moving towards the poles, while others were still on the aequa- 
tor. In anaphases, viewed from the side, very small chromosomes were 
frequently seen on the spindles, while larger ones were present in the 
groups near the poles. 
To me it does not appear to be improbable that the chromosomes 
which had formed gemini in the heterotype division, proceed to split 
in the homotype division and that subsequently the smaller chromoso- 
mes, arisen by fission of univalent ones, split again in the homotype 
division, or possibly, are segmented as a consequence of which very 
small chromosomes arise which reach the poles some time after the 
other chromosomes have reached them already. Possibly not all chro- 
mosomes arisen by fission of univalent ones, split again in the homo- 
type division, because in this division the number of chromosomes 
lagging behind was always smaller than in the heterotype divison. 
They were probably distributed unsplit over the poles, according to 
chance. 
Occasionally small quantities of chromatine could still be observed 
on the spindles, after the new cellwalls in the spindles and the daugh- 
ternuclei also had already been formed. A few chromosomes consequent- 
ly got probably lost in the cytoplasm. In the same preparations of 
the inflorescence, in which such an unusual large number of unival- 
ent chromosomes was present in the nuclei of the pollen-mother- 
cells, the division could—although not in the majority of cases — 
take place in a still more irregular way. Dyads namely did occur, 
which showed a large nucleus in the one and a small nucleus in the 
other cell. Many dyads were even met with in which the one cell 
contained a very large nucleus, the other none at all. Such a case 


