AND SPECIES-HYBRIDS WITHIN THE GENUS SACCHARUM 297 
ling of two chromosomes in the PMC and EMC of the latter to pair, 
as a consequence of which 3 chromosome-pairs and 2 unpaired chromo- 
somes would have been present in those mothercells. Because occa- 
sionally the two unpaired chromosomes could have travelled towards 
one pole, one of the daughternuclei would have obtained 5, the other 
3 chromosomes. The result would have been the formation of pollen- 
grains and embryosacs with 3 and such with 5 chromosomes, besides 
the normal ones with 4 chromosomes. In case of fertilisation of an 
eggcell with 3 chromosomes by a spermnucleus with 3 chromosomes, 
a daughter-individual would arise, with 3 chromosomes in the haploid 
phase. In the case of Crepis Reuteriana, a species with 3 chromosomes, 
such irregularities often occur. He writes: „Solche Unregelmässigkei- 
ten sind nun keineswegs eine Seltenheit bei C. Reuteriana. Von 60 in 
bezug auf Chromosomenzahl untersuchten Kernen waren fünfzehn 4- 
resp. 2-chromosomig, also etwa 30 % (25% was probably meant). 
Auch unter den 4-chromosomigen Arten habe ich solche Unregel- 
massigkeiten, wenn auch selten, gefunden.” 
Among the investigated forms of S. spontaneum small irregularities 
in the division were but exceptionally found. To explain the origin of 
S. officinarum or of Chunnee from S. spontaneum great irregularities 
in the reduction division of the last species would have to be imagined, 
leading to the obtention of very different chromosomenumbers by the 
gametes. Such irregularities would probably cause the sterility of a 
considerable part of the pollengrains. S. spontaneum however is, in its 
numerous forms, according to BARBER also, a very fertile species, so 
that the occurence of such large irregularities is not probable. Abstrac- 
tedly it would not be impossible that gametes with a diminished chro- 
mosomenumber could be fertile in certain cases, so that they, be it but 
rarely, could give rise to individuals with a much smaller chromosome- 
number. To assume however, that they would acquire by such a loss 
or, simultaneously with it by some other cause, new properties, among 
which, in our case , the formation of sugar would be one of the most 
important, is highly improbable indeed. Moreover, different forms, so 
arisen, would doubtless possess very different chromosomenumbers, 
while Chunnee and Ruckree II have a similar or even equal chromosome 
number and S. officinarum in the case of 5 very different forms, pos- 
sesses in the haploid phase a number of exactly 16 chromosomes less, 
than that of S. spontaneum. 
