Mutations and Evolution. 
147 
distinct types, one with rather broadly linear petals and rather 
thick flower-buds, the other with narrowly linear petals and slender 
flower-buds. A third cruciate type grown at Amsterdam in 1903, 
from seeds collected at Jaffrey, New Hampshire, differed in 
having a much longer calyx tube, and a more slender, less nutating 
stem. MacDougal (1905) gave descriptions of a third type from 
the Lake George seeds, which he identified with the New 
Hampshire form, but this identification is doubtful, as Bartlett points 
out. None of these three types is the same as E. cruciata Nutt, 
which came from Massachusetts. The two Lake George forms were 
described (Bartlett, 1914) as E. atrovirens Shull and Bartlett and 
E. venosa Shull and Bartlett. These may of course be cruciate 
varieties of the local species where they occur, but it seems more 
probable that they represent independent species. This would 
indicate that they have undergone further evolution since the 
cruciate mutation took place, or that the parent broad-petalled 
type has since become extinct. 
Herbarium specimens with cruciate flowers, which have been 
placed under E. cruciata Nutt., but no doubt belong to several 
distinct types, have been recorded (Bartlett, 1914), from Nova 
Scotia (Sable Island), Maine, New Hampshire, Vermont, New 
York and Massachusetts. Another cruciate type with small 
flowers is represented by a specimen in the British Museum 1 
collected in British Columbia in 1909—evidently an independent 
local variety or species. Bartlett also grew 50 plants from seeds 
of one individual collected at Springfield, Missouri. They were 
uniform with the exception of one plant, a single branch of which 
bore cruciate flowers. The petals were narrowly oblong, rather 
than linear. This suggests previous crossing with a cruciate 
mutant, which may very well have occurred in the field. Bartlett 
has cultivated a cruciate CEnothera from Mobile, Alabama, and 
describes another one, (E. stenomeres, from Montgomery, Co. 
Maryland, which has given rise to mutations. Yet another cruciate 
species has been cultivated by Shull from Long Island, and has 
been described as E. cleistantha Shull and Bartl. (Bartlett, 1915e). 
It is not closely allied to any other known species, and the origin 
of the cruciate feature is therefore probably of ancient date. The 
flowers as a rule never open, and the most striking features of the 
species are its extreme leafiness and dense branching. It is 
obvious that the possibilities of natural crossing in this species are 
1 Gates, 1915a, p. 21. 
