300 A CYTOLOGICAL INVESTIGATION OF SOME SPECIES 
Paarung während der Prophase stattgefunden hat, die aber später 
zurückgeht, so dass die Chromosomen schon während der Diakinese 
sich trennen und als ungepaarte Chromosomen in die Spindelfigur ein- 
gereiht werden.” 
; Chromosomes behaving in a similar way were found by ROSENBERG 
‘in dividing EMC of H. pilosella, a species also containing 36 chromoso- 
mes in the somatic cells. 
Hybrids between A. excellens, itself forming gametes with different 
chromosomenumbers and H. aurantiacum or H. pilosella showed very 
different chromosomenumbers. which shows that gametes with diffe- 
rent chromosomenumbers formed by the parents are fertile. It is not 
known whether gametes of Saccharum officinarum with chromosome- 
numbers deviating from 40 can be fertile, in view of the fact however, 
that gametes of Hieracium-species with different chromosomenumbers 
can be fertile we have to count with this possibility in the case of Sac- 
charum officinarum also. ¢ 
During his investigation of the fitness of pollen of sugarcane, JES- 
WIET found that the inflorescences of Bandjermasin hitam furnished 
but very little pollen; part of the anthers remained closed. Of the 
pollengrains 50 % contained starch and had a normal appearance; the 
other pollengrains had remained very small and often had retained the 
shape of the cells of the tetrads, and were empty. 
In the case of Ardjoeno he found, on the contrary, that all anthers ope- 
ned and liberated a large quantity of pollen, 80 a 90 % of which con- 
tained starch. 
After selfing, Ardjoeno always gives a large progeny, Bandjermasin 
hitam a very small one. Investigation of the mikrosporogenesis of 
Ardjoeno reveals the presence of but little incomplete reduction, while 
such incomplete reduction is found almost exclusively in the case of 
Bandjermasin hitam. It is possible that this incomplete reduction in 
the case of Bandjermasin hitam, causes the origin of a great many pol- 
lengrains with combinations of chromosomes which cause the death of 
the gametophyte or impeach the fertility of the male generative nuclei, 
while this happens to a far less degree in the case of Ardjoeno with its 
usually much more regular reduction-division. In as much, very pos- 
sibly other factors, causing pollensterility may exist, an investigation 
of a large number of sugarcane-kinds, extended over a number of 
years will be required before it will be possible to conclude, to the. 
