555 
Gates .— The Trisomic Mutations of Oenothera. 
iana in the garden of deVries and identified by him as typical rubrinervis , 
cytological examination of the root tips showed a small extra chromosome, 
somewhat larger than the small one in aberrans. It is significant that this 
mutant also arose from lata x Lamarckiana. While the origin of this 
chromosome fragment is therefore clear, its significance is less certain. Mut. 
rubrinervis ^ de Vries, is recorded as a form which breeds true and has no 
marked pollen sterility (i. e. the usual 50 per cent). It is therefore not clear 
whether this fragment will prove to be a constant feature of the type which 
de Vries calls rubrinervis. It has long been evident that the form identified 
in American cultures as rubrinervis differs in certain features from the type 
of de Vries. It is a taller, stronger plant without a zigzag stem but with 
the same red stripes on the sepals, and appears to be the same as the deriva¬ 
tive from rubrinervis crosses which de Vries ( 1913 , p. 192) has since called 
subrobusta. The latter may in turn give rise to rubrinervis (according to 
de Vries) when self-pollinated. The taller, stronger plant is the one which 
has been called rubrinervis in all of my experiments. It certainly has 14 
chromosomes without any extra fragment (Gates, 1908 ). There would 
appear to be still some uncertainty regarding the unity of the forms which 
de Vries has called rubrinervis. 
Finally, brief reference may be made to mut .lasiopetala, which originated 
from Oe. stenomeres in the cultures of Bartlett ( 1915 ). It has hairy petals 
and is very late in flowering, forming terminal rosettes on its branches. 
Less than half the pollen appears viable, and as its offspring are dimorphic, 
like those of other trisomic mutants, it probably also has 15 chromosomes. 
The hairy petals are a new character in the genus and it is totally different 
from any other trisomic form. The tetraploid mutant gigas from Oe. sieno- 
meres also gave rise to a secondary mutation lanosa with hairy petals like 
lasiopetala and also the ‘ filaments of the stamens alternate with the petals 
were densely lanose ’. It is therefore parallel with lasiopetala but more 
extreme. Possibly this form had 29 chromosomes. 
We may next consider the occurrence of plants with 16 chromosomes, 
either as mutations or in the offspring of simple trisomic forms. Their 
appearance is rare, and this confirms the fact that 8-chromosome pollen 
grains rarely function or rarely retain their full chromosome complex in the 
male nuclei. Miss Lutz ( 1917 ) obtained two lata -like mutants from separate 
cultures of Lamarckiana in 1908 and 1910. Both had 16 chromosomes, 
but they were ‘ in no sense identical forms ’. Like lata, they had crinkled 
leaves, one broad, the other narrower ; both had yellow-green foliage, irregu¬ 
larly shaped buds, and were male sterile. A third form with 16 chromosomes 
was a dwarf arising from lata selfed and having broad leaves in the rosette 
stage. In the experiments of van Overeem ( 1920 ) several forms with 16 
chromosomes (double trisomic) have appeared. In the offspring of Oe. 
Lamarckiana scmigigas the chromosome number ranged from 34 to 28 and 
