294 A CYTOLOGICAL INVESTIGATION OF SOME SPECIES 
one or the other species of Saccharum, cytological investigation may 
therefore be able to give a decisive answer. This may also be the case 
with forms, which, after selfings give a heterogeneous progeny 1). 
It was already pointed out in the introduction, that the chromosome 
numbers frequently form a séries within a genus. The chromosomenum- 
bers 40 in the case of S. officinarum and 56 in that of S. spontaneum 
could be considered as terms of a series with the basic number 8. 
It would however be very risky, to say now already, that sucha 
series of chromosomenumbers occurs within the genus Saccharum. 
The diploid chromosomenumber of Chunnee is 90 à 91, and may cer- 
tainly not be fixed at 96. The haploid number therefore is not 48, but 
approximately 46. The chromosomenumber of Chunnee therefore does 
not form part of a series of 8. This however does not prove the non- 
existence of such a series, because one could presume that ancestors of 
Chunnee and Ruckree IT did possess 48 as the haploid and 96 as the 
diploid number, but that irregularities in chromosome-distribution of 
these ancestors had Jed to the formation of gametes with a chromoso- 
menumber slightly inferior to 48. Chromosomes which remained un- 
paired and which during reduction were distributed irregularly over 
the daughternuclei or all of which had not become included in the pol- 
lengrains, could have played a role. 
A comparison of the size of the nuclei of PMC of the investigated 
Saccharum-forms during the diakinesis-stage, shows that the nuclei of 
S. spontaneum are larger than those of Chunnee, while those of S. offi- 
cinarum are smaller. ; | 
In the case of the banana-races: Dole, Radjah Siam and Kladi 
TISCHLER ?) had found, that the nuclear volumina of the PMC in sy- 
napsis stood to one another in the proportion 1 : 2 : 3. The chro- 
mosomenumbers of these in the gametophyte were 8 : 16 : 24, so that 
the nuclear volume is proportional to the chromosomenumber. 
Barry %) found, that the nuclear volumina of Triticum vulgare and 
1) It results from a preliminary investigation, that Teboe Hitam Rokan, which 
in habit differs somewhat from Saccharum officinarum, possesses about 30 chro- 
mosomes in the haploid phase, while Loether’s cane, which, on selfing, gives types, 
differing from S. officinarum, has about 50 chromosomes in the haploid phase. 
*) G. TiscHLER. Untersuchungen ueber die Entwicklung des Bananenpollens 
I. Archiv f. Zellforschung. Bd. V. p. 649. 
3) W. Barry. Die Godronschen Bastarde zwischen Aegilops- und Triticum- 
arten. Zeitschr. f. ind. Abst. u. Vererbungslehre. Bd. XX, 1919. 
