AND SPECIES-HYBRIDS WITHIN THE GENUS SACCHARUM 301 
existence or non-existence of a connexion between the mode of reduc- 
tion and fertility, the more so, as the degree of fertility of the pollen 
differs in different years in the case of some kinds of sugarcane. 
Green German New Guinea gives, upon selfing, a large number of 
aberrant types besides the normal ones, types with very short inter- 
nodes. One could put the question whether these aberrant types are 
the product of unions between gametes with aberrant chromosome- 
number. This would require the investigation of a large of number of 
plants of normal and abnormal habit. Many of these plants would 
certainly not flower and therefore would have to be discarded in the 
- investigation, as an investigation of somatic nuclei gives no satisfac- 
tory results. Such an investigation would consequently require much 
time, while it would be doubtfull whether the desired end could be 
reached. 
It may happen, as an exception, in the case of Saccharum officina- 
rum var. Green German New Guinea that almost all chromosomes fail 
to pair in the prophase of the reduction division. During diakinesis the 
chromosomes are then, generally, arranged two by two, often lying 
parallel to one another, but they are not united with one another. As a 
consequence they lie in the metaphase, approximately in diploid num- 
ber, in the nuclear plate, and are subsequently, without having split 
longitudinally, distributed, according to chance, over the poles of the 
spindle; a very irregular tetrad-formation results. 
A similar division-process was found by ROSENBERG in: the PMC of 
Hieracium laevigatum and H. lacerum. These species have 27 chromoso- 
mes in the somatic cells. Here also it may hapen that all chromosomes 
remain unpaired in the prophase of the division of the PMC. In the 
metaphase 27 chromosomes also are then found which, without fission, 
are distributed, according to chance, over the poles. The number of 
chromosomes at the two poles was sometimes approxiately equal, some- 
times very different. A second division did not always follow; fre- 
quently the diad-cells themselves were transformed to pollengrains. 
This kind of division has been named , halbheterotypische Teilung’ by 
ROSENBERG. The abnormal division found by me in the case of Saccha- 
rum officinarum var. Green German New Guinea forms an analogon to 
this division, with the exception that in the prophase of the division 
of G. G. N. G., the chromosomes are arranged in sets of two. 
