266 BOTANICAL GAZETTE [OCTOBER 
Lamarckiana in 1913, and got from the seeds a uniform generation 
of 60 flowering individuals, all of which proved to be Lamarckiana. 
I conclude from this fact that the pollen of the pallescens 
plants does not transmit the characters of the race, exactly 
as in O. scintillans and O. cana. 
O. Lamarckiana lata mut. Lactuca (fig. 5, C).—In the summer 
of 1913 I found, in a race of O. lata which had been fertilized in the 
previous generations (1905 
and 1907) by O. Lamarck- 
iana, a weak plant which 
seemed to be new to me, 
but showed evident signs 
of affinity with the incon- 
stant types of O. cana and 
O. pallescens as previously 
described. It was fertil- 
ized, therefore, purely by 
its own pollen. It yielded 
0.8 cc. of seeds, which 
were sown in 1914 and 
gave rise to 65 plants, one 
of which was a mutant 0 
the ordinary type of O. 
nanella, and subjected to 
the same bacterial disease 
Fic. 5.—Typical radical leaves of A, Oeno- 
thera Lamarckiana mut. liquida; B, mut. cana; : 
C, mut. Lactuca; June 1914. which so often deforms the 
dwarfs of my race. Among 
the others, two types were represented in about equal numbers. One 
type was exactly like normal Lamarckiana; it counted 36 individuals, 
almost all of which have flowered, without showing any recognizable 
difference from the original wild species. The remaining 28 con- 
stituted a new and uniform type, repeating the characters of the 
parent plant of 1913, so far as these had been noticed and recorded. 
At the time of planting out, in the beginning of May, they very much 
resembled the compact rosettes of O. nanella, but without any signs 
of the disease. About the middle of June, when the rosettes of 
the type of Lamarckiana were growing very fast, those of the new 
