Gates .— The Chromosomes of a Triploid Oenothera Hybrid. 567 
Fig. 4 shows another homotypic anaphase in which more chromosomes 
have been lost, for there are only 8 pairs of separating chromosomes on the 
left-hand spindle and 7 pairs on the right. Hence the remaining 6 whole 
chromosomes must have been left out of the daughter nuclei on the hetero¬ 
typic spindle. 
In Fig. 5 a homotypic spindle is in telophase, with 7 chromosomes at 
either pole and an eighth divided chromosome remaining behind. Only 
8 chromosomes reached the daughter nucleus, which is here dividing. As 
the number of chromosomes on the other homotypic spindle is unknown, it 
is impossible to say how many of the other 13 chromosomes reached the 
opposite pole of the heterotypic spindle. Fig. 6 shows a homotypic spindle in 
which 8 chromosomes have reached either pole and a ninth split chromosome 
is disintegrating on the spindle. Probably in this mother-cell at least one 
chromosome was lost from this end of the heterotypic spindle. F'ig. 7 
represents a not infrequent condition in which only 7 chromosomes have 
arrived at the poles of a homotypic spindle, the eighth disintegrating. 
These figures show that a varying number of chromosomes may be lost 
during either of the reduction divisions in this triploid hybrid. Occasionally 
all the chromosomes may reach the daughter nuclei in the heterotypic mito¬ 
sis, but usually one, two, or more unpaired chromosomes are left behind to 
disintegrate. All the chromosomes which reach the daughter nuclei in the 
heterotypic division apparently undergo a split. When the homotypic 
spindles are formed and these split chromosomes separate, a certain number 
of them may again be left behind, so that the number of chromosomes 
ultimately reaching the four daughter nuclei is usually only 7 or 8. 
In an early paper (Gates, 1909 ) on Oenothera lata xgigas I showed how 
the 21 chromosomes usually separate 10-11 on the heterotypic spindle, but 
sometimes 9-12. 1 Apparently in that material all the chromosomes usually 
reached the two daughter nuclei. The chromosome numbers in interkinesis 
therefore ranged only from 9 to 12, and this entirely owing to the sometimes 
irregular chromosome distributions. In the present hybrid Oenothera rnbri- 
calyx xgigas, however, there is evidence that from one to six chromosomes 
are frequently lost on the heterotypic spindle, since the numbers on the 
homotypic spindles are frequently only 8, 9, or 10. 
That similar conditions occur in the pollen formation of other triploid 
Oenotheras is indicated by recent results of van Overeem ( 1920 ), to which 
brief reference may be made. From Oe. Lamarckiana mut. semigigas (21 
chromosomes) crossed with Lamarckiana he obtained plants as follows: 
1 x j 4, 9 x j 5, 4 x 16, 2 x i 7, 2x18, 1 x 20, and 1x21 chromosomes ; from 
semigigas xgigas x x 21,2 x 22, 12 x 23, 21 x 24, 25 x 25, 19 x 26, 6 x 27, and 
1 x 28. Oe. biennis semigigas x Lamarckiana gigas gave corresponding 
1 These results have recently been confirmed by van Overeem ( 1922 ) in an important paper 
dealing with various triploid and other forms. 
