19141 
GATES—SOME (ENOTHERAS FROM CHESHIRE AND LANCASHIRE 393 
two on the right in pl. 22 fig. 18 were greatly overgrown and 
were far larger than ever appear even in Œ. mut. gigas. These 
forms have not been sufficiently studied since to give an ade- 
quate account of them. 
It will be obvious that the forms described here under the 
names multiflora, multiflora elliptica, rubrinervoides, tardiflora 
and rubritincta are not pure species or even true-breeding races. 
They are undoubtedly as diverse from each other as average 
species, however, and many systematie species if bred experi- 
mentally would probably not breed true within narrower limits 
than these races have done. One feature of interest attaching 
to these races is the fact that the main type persists essentially 
unchanged, though various mutants and heterozygous forms 
are thrown off. The behavior is not, in the main, like the Men- 
delian process of recombination. Repeated selfing of each race 
usually decreases its variability by eliminating various hybrid 
elements. But this process does not extend to the basal differ- 
ences between the races, which, as we have seen, remain as 
unlike as they were before. In this aspect the hereditary 
behavior of these races resembles that of Œ. Lamarckiana. 
But there are a number of differences which I need not fully 
consider. Thus (E. multiflora gives rise to its variety elliptica 
much as though it were split off from a heterozygous condi- 
tion, and the variability of rubritincta in leaf-width, as well as 
its production of numerous dwarfs, is unlike anything in the 
behavior of (Е. Lamarckiana. 
Many other equally distinct types were derived from this 
locality (see, e. g., pl. 22 figs. 18, 19), but they have not been 
cultivated in subsequent generations. 
Œ. LAMARCKIANA FROM ST. ANNE'S 
Іп 1910 I obtained seeds from a colony of Œ. Lamarckiana 
growing by the Manchester Children's Hospital Convalescent 
Home, at St. Anne’s-on-Sea. Many of these were found in 
later cultures to agree exactly with the Lamarckiana of de Vries 
except in the red color pattern of the sepals. I was formerly in- 
clined to lay little stress on this difference but there is no doubt 
that it is inherited. Тһе fact therefore remains that a precise 
duplicate for de Vries's race of Œ. Lamarckiana is relatively 
