545 
Gates .— The Trisomic Mtitations of Oenothera . 
a megaspore mother-cell, only one functional megaspore with eight chromo¬ 
somes would have resulted. The occurrence of any aberrant form in a pair 
of individuals therefore suggests at once that they have probably arisen 
from an irregular chromatin distribution in the heterotypic mitosis of 
a pollen mother-cell of the previous generation. In this case the non-dis¬ 
junction of a pair of chromosomes must then have occurred in a pollen 
mother-cell of Oe. Heivettii. 
We may now consider briefly the main cytological features of this 
aberrant type, confining ourselves to diakinesis and the heterotypic mitosis. 
Figs, i and 2 , PI. XI, represent late stages in diakinesis. In the former 15 
chromosomes can clearly be counted. It will be seen that nearly all of the 
chromosomes, but not all, are arranged in pairs, these pairs in many cases 
forming complete rings. It has long been known that in Oenotheras of the 
Lamarckiana and biennis group the formation of ring chromosomes seldom 
occurs (Gates, 1908 ; Davis, 1910 ), while Davis ( 1909 ) has described the con¬ 
stant formation of ring chromosomes during diakinesis in Oe. grandiflora , 
and Cleland ( 1922 ) has recently figured the regular occurrence in Oe. fran- 
ciscana of five pairs of chromosomes in rings, the other four forming a single 
large ring. In Figs. 1 and 2, referred to above, there are five pairs of ring 
chromosomes in each case. In no case have more than five ring-pairs 
been found, although there are sometimes less, and usually not more than 
one ring persists on the heterotypic spindle. Unfortunately no material of 
the parent species Oe. Heivettii at this stage has been examined, so it is 
unknown whether this species is like Oe, grandiflora in having only ring 
chromosomes, but it appears more probable that it may be like Oe. francis- 
cana , having five ring-pairs and four other chromosomes more loosely united 
into a single ring. 
As previously pointed out by the author, the absence of a clear pairing 
of all the chromosomes during diakinesis and the heterotypic mitosis is an 
indication of a weak attraction between the chromosomes constituting 
a pair. This is the characteristic condition in Oe. Lainarckiana and its 
derivatives. Davis and Cleland have argued that the occurrence of ring 
chromosomes in Oe. grandiflora and Oe. franciscana 1 is a sign of a ‘ pure 
species \ In the present instance, however, we have as many as five ring- 
pairs of chromosomes in a 15-chromosome mutant occurring in an F 1 hybrid 
between forms belonging to quite distinct species. One of these rings might 
of course be descended from the pair which failed to separate in the Hewettii 
pollen mother-cell, but there is no evidence that such a pair will persist in this 
1 It should be pointed out that the writer made a study of the variability of Oe. franciscana , 
which Cleland stresses as a pure species. While in California in 1916 a study was made of large 
wild colonies growing near San Francisco, and also of the variations of this species in cultures. The 
wild variations were particularly conspicuous in size of flower, pigmentation, and the number of red 
papillae on the sepals. It was evidently a population of interbreeding forms and must have been 
heterozygous for a number of characters. 
O O 2 
