554 Gates.—The Trisomic Mutations of Oenothera. 
consequent relationships with each other, with the intention, if possible, of 
bringing order out of what appears to be chaos. But it seems that, with 
some exceptions, any trisomic form is capable of giving rise to any other. 
The experiments of de Vries \\ ith cana and other forms lead him to the con¬ 
clusion that none of these trisomic mutants transmit their characters through 
the pollen. This is in accord with the cytological studies of lata (Gates and 
Thomas, 1914 ), in which it was shown that very few pollen grains can be 
formed having eight chromosomes. But in trisomic mutants such as the one 
described in this paper, where a larger percentage of functional pollen is 
producedand thepollen nuclei not infrequently receive eight chromosomes, it 
is probable that the extra chromosome may often be lost in the later nuclear 
divisions of the male gametophyte. When used as the female parent in 
crosses, these trisomic forms frequently appear in numbers approaching 50 
per cent, of the offspring. From this fact it seems clear that in megaspore 
formation the chances are equal that the extra chromosome will pass to 
either pole, the defect in numbers of the trisomic form being the result of 
a lesser viability of gametes with the extra chromosome, or of the extra 
chromosome sometimes being left behind on the spindle. There is, on the 
contrary, some evidence of greater viability on the part of zygotes with 
an extra chromosome (at least in the case of lata). 
We must now refer to certain other interesting forms in Table I (p. 548) 
which have not yet been mentioned. These are aberrans, Lutz, and rubri- 
nervis , de Vries. Miss Lutz ( 1916 ) describes aberrans as a mutant from lata 
x Lamarckiana, two individuals occurring from this cross in 1908 and 1909. 
The type differs from Lamar ckiana by no very marked characters except 
the almost complete pollen sterility. A moderate abundance of pollen was 
produced, but containing few good grains, and it was found almost im¬ 
possible to obtain seed by self-pollination. The plants had more slender, 
tapering buds than Lamar ckiana, but produced an abundance of flowers 
with slightly lighter yellow petals. Cytological examination of root tips 
showed the constant presence 1 in both plants of a small chromosome frag¬ 
ment in addition to the usual 14 (hence I4 +1 ). As Miss Lutz points out, 
this type no doubt arose through a fragment of a disintegrating chromo¬ 
some becoming included in one of the daughter nuclei, a condition which 
was shown actually to occur (Gates and Thomas, 1914 ) in the meiotic pollen 
divisions of Oe. lata forms. The inclusion of such a fragment has been 
sufficient to alter appreciably the characters of the plant; the fragment has 
regularly divided in mitosis, but its presence during the meiotic division was 
a sufficient disturbance to cause almost complete pollen sterility. 
The other fact recorded by Miss Lutz ( 1916 ) which is germane to the 
present discussion is that in a rubrinervis mutant arising from lata x Lamarck- 
1 Occasionally the small chromosome could not be found in cells in which it apparently could 
not have been concealed. 
