290 A CYTOLOGICAL INVESTIGATION OF SOME SPECIES 
90 or 91, 91 being by far the more- probable number however. When, 
therefore, the pairing in the PMC is complete, this would probably lead 
to 45 gemini and one single chromosome. 
It seems impossible to admit 91 as the diploid chromosomenumber of 
Chunnee, as diploid chromosomenumbers are usually pair. The possi- 
bility, to my way of looking at the question, is however not excluded 
that Chunnee and Ruckree are descendants of a Saccharum-species with 
48 chromosomes in the haploid and 96 in the diploid phase. By irregu- 
lar reduction-division of such a species, gametes could have been for- 
med with a little less than 48 chromosomes, and these, after union, 
could have produced a form with 91 chromosomes in the diploid phase. 
We shall again refer to this point in the third chapter. 
Counts of chromosomes in the homotype division of Chunnee are 
not possible. 

Fig. 84. 
Fig. 83. Pollen-mothercell of Ruckree IT in diakinesis x 2300. 
Fig. 84. Nuclear plate of the heterotype division of Ruckree II x 2300. 
The preparations of Ruckree IT cannot be distinguished from those 
of Chunnee, the dividing PMC of both are quite similar. 
Of 10 counts of diakinesis-nuclei of Ruckree II the results vacillated 
between 46 and 48. 
Fig. 83 shows a diakinesis-nucleus containing probably 45 or 46 
chromatine-elements. 
