438 BOTANICAL GAZETTE [DECEMBER 
(This plant is shown in figs. 5 and 14.) All three mutations are 
discussed here, but full data and illustrations are reserved for a 
paper entitled ‘‘Certain mutations and hybrids of O6enothera 
pratincola,”’ to appear later in this journal. 
Mut. angustifolia—It has been found that mut. angustifolia 
from ordinary strains gives no descendants resembling itself; aside 
from the usual small proportion of other mutational types, the 
progeny consisting of f. ¢ypica only. The variation, although 
striking and entirely discontinuous, appears to be somatic. In this 
connection it is interesting to note that a perfect branch of f. typica 
has been observed as a bud sport on mut. angustifolia. As would 
be expected, mut. angustifolia crossed reciprocally with f. typica 
gives f. typica together with the usual few mutations. The 
hereditary behavior may be stated: 
mut. angustifoliaX mut. angustifolia > {. typica 
mut. angustifoliaX{. typica > f. typica 
f. typicaX mut. angustifolia > f. typica 
The behavior of mut. angustifolia from Lexington E is most 
remarkable. The parent plant was a sister of the typica plant whose 
progeny is analyzed in table III, but the degree of mutability proved 
to be much more extreme than in the case of the typica sister. 
Only two plants in the progeny, out of a total of 505, were f. typica; 
the other 503 plants, 99.6 per cent of the progeny, were mutations 
belonging to the group characteristic of Lexington E. The results 
are summarized in table IV. 
The cross mut. angustifoliaX{. typica and the reciprocal gave 
respectively 100 per cent and 91.5 per cent of mutations. By com- 
parison with table III it will be seen that each of the reciprocal 
crosses tends to show the same degree of mutability as the female 
parent. To be sure, there would seem to be a considerable dis- 
crepancy between 74.3 per cent, representing the mutability of 
f. typica, and 91.5 per cent, representing the mutability of f. typica 
X mut. angustifolia. No significance can be urged for this dis- 
ctepancy, however, when we consider that one of the cultures from 
a single capsule of f. typica contained g2.2 per cent of mutations 
among 102 plants, as compared with 91.5 per cent of mutations 
