IOWA ACADEMY OP SCIENCE 
117 
pairs when the corolla expands and then rolled inwards.” The corolla 
is described as one and a half to nearly two inches in diameter. 
From this description and the careful measurements it is evident that 
this plant had small flowers about the size of the American 0. biennis. 
The synonymy and other statements', which were copied from book to 
book, cannot be taken as meaning anything in the present connection. 
Contemporaneously (1806) Sowerby, as we have seen, pictures a large- 
flowered form closely resembling 0. Laniarckiana, under the name 0. 
biennis j so that it is quite evident that at this time no distinction was 
drawn between 0. biennis and 0. Lamarckiana forms, although 0. grand- 
iflora had been segregated. 
The condition of the plants now growing wild and freely intercrossing 
on the sand-dunes near Liverpool, is probably somewhat similar to what 
it was in their original home in Virginia, although it is probable that 
in their original habitat the individuals were much more scattered, owing 
to the nature of the habitat and the competition of other plants. For 
this reason, crosses between the different species were much less likely 
to occur, but that such crosses did occasionally occur there can be no 
doubt. It seems characteristic of species which' have 'become “weeds” 
in another country, that they grow in large numbers of individuals 
closely aggregated in localized areas, while in their native habitat they 
are more uniformly scattered over larger areas, taking their part in the 
regular flora of the country. The reasons for this difference in distribu- 
tion I shall not discuss here. In the case of the open-pollinated Evening 
Primroses, it is not at all improbable and indeed may be regarded as 
certain, that crosses between different forms did occasionally occur where 
their ranges of distribution overlapped. In the case of the three species 
we are considering here, it is probable that before the white man’s inva- 
sion of the continent, all three were to be found over a large part of the 
country. Since then the small-'flowered, close-pollinated 0. biennis and 
its related forms, such as O. Oakesiana, 0. muricata and 0. fruticosa, 
have continued to maintain themselves, while the open-pollinated 0. 
grandiflora seems to have nearly or quite disappeared from its Eastern 
range in Virginia and Carolina, and 0. Lamarckiana seems to have be- 
come cjuite extinct on this continent. 
It would seem, therefore, that the close-pollinated species have been 
more successful in their competition with the conditions introduced by 
civilization, than the open-pollinated forms. This might be expected, 
because in close-pollinated forms seed production is always certain to 
follow flowering, while in open-pollinated species, with increased enemies 
