go BOTANICAL GAZETTE [FEBRUARY 
Lexington A, B, and E germinated readily. Since no variation 
was noticed among the hundreds of seedlings of these three strains, 
all were discarded except 30 of each, which were potted off for the 
field cultures. The seeds of Lexington C, however, had been 
planted too deeply in unsuitable clay soil, and, although the seed 
LOXS/ os LOs7 
.—F, progeny of Oenothera et en four of the 9 seedlings which consti- 
nas ‘i first culture of Lexington t upper plants, Lexington C-1 and C-2, are 
mut. nummularia; the latter bore seed Ww hich gave rise to the F, culture he to in 
table I; the 2 lower plants are typical O. pratincola. 
pan received the same treatment as the rest, weeks passed before 
any seedlings appeared. At length 9 plants were obtained which 
were potted off. Almost from the first, they showed remarkable 
variation among themselves. Six (nos. 3, 4, 5, 6, 8, and g) were 
typical O. pratincola, and agreed in all characters with the seed- 
