The Stature and Chroinosomes of Oenothera gigas, De Vries. 527 
trast to the condition in other genera of plants, where the ckromo- 
somes are all regularlv paired. Howevev, in Hieracium (Bosexberg 
1907) and probably also in Galtonia and Tradescantia (Miyake 1905), 
as I have suggested elsewhere (Gates 1908 b, p. 23), a similar failure 
to pair is often exbibited. These cases appear to be exceptions to 
tbe general law enunciated by Montgomery in 1901 from bis ob- 
servations on Hemiptera, tbat homologous chroinosomes of maternal 
and paternal origin pair witb eacb otber in synapsis. Later obser- 
vations on a variety of forms in wbich there are morphological ckro- 
mosome differences show tbat ordinarily chromosomes of similar size 
and sbape pair with eaek other, and tliis seems amply to justify tlie 
view of Montgomery, wbich bas been widely adopted. It is of 
course conceivable tbat in such cases as Oenothera where there is 
a partial failure of pairing, some invisible forces on the keterotypic 
spindle may yet determine tbat homologous maternal and paternal 
chromosomes shall enter opposite nuclei. But this seems rather un- 
likely. It is moreover, contrary to fact in the case of the hybrid 
0. lata X 0. gigas, in which, as 1 have shown (Gates 1908 a) 1 ), the 
chromosome distribution at tbe time of reduction bears no relation 
to the parental chromosome numbers, but on the contrary they are 
usually separated into two equal groups as nearly as an odd number 
(21) can be. The regularity of this Segregation into groups of 10 
and 11 chromosomes, with only occasional cases of 12 and 9, and 
no observed cases of greater irregularity, makes it probable that 
some regulative factor is present, which determines this distribution, 
although what it may be is not obvious. 
In fig. 3 the full number, 28 chromosomes, are in view. Two 
chromosomes are evidently paired on the equatorial plate of the 
spindle, and each shows the longitudinal split for the next mitosis, 
showing that in some cases at least, the split for the second mitosis 
is longitudinal. A certain amount of Variation in the size and shape 
of the chromosomes is observed, but this is to be expected from 
tkeir viscous or semi-fluid consistency. No constant morphological 
chromosome differences have been found. They are apparently 
eitker less common or less easy of detection in plants than in ani- 
mals, perhaps owing to greater variability in shape. Fig. 4 is a 
polar view of the central region of the keterotypic spindle, the chro- 
mosomes not being in one plane as represented but only coming into 
*) A paper dealing with this hybrid is now in press in the Botanical Gazette. 
35 * 
