Recent Papers on CEnothera Mutations. 297 
Professor De Vries (13) with Mr. Bartlett made a visit to the 
type-locality for 0. grandiflora, in Alabama, last year—the locality 
where Bartram discovered this species in 1778. They found both 
O. grandiflora and 0. Tracyi, and a number of races or variations of 
each. These forms freely intercrossed, at least in the case of 
grandiflora , so that no forms were free from the suspicion of 
hybridization. This confirms a fact which had already been 
emphasized by Gates (15), namely, that in open-pollinated plants 
the intercrossing of many closely-related races which make up the 
population is continually taking place in every generation. The fact 
that O. Lamarckiana has undergone crossing does not therefore 
render its condition essentially different from that of any other 
allogamous species. 
Much attention has been devoted to the history not only of O. 
Lamarckiana, but of all the related species in Europe. Many of the 
early records, from 1614 to the present time, have been brought 
together by Gates (14), and in this and subsequent papers many of 
the plants referred to or figured under pre-Linnsean polynomials 
have been identified with reasonable certainty. The writer’s 
first conclusion, that the original form brought to Europe most 
resembled 0. Lamarckiana, has since been modified, but it seems 
certain that forms agreeing with 0. Lamarckiana were growing wild 
in Europe at the time this species was recognized and described 
from garden material in Paris about 1797. 
In drawing any conclusions concerning the origin and history 
of 0. Lamarckiana, whether through crossing or otherwise, the 
following crucial facts must be kept in mind : (i.) That a large- 
flowered CEnolhera from “Virginia” was recognized by Ray as early 
as 1686. This must have belonged to a race either of 0.grandiflora 
Solander or 0. Lamarckiana Ser., and the possibility that it may 
have been the latter is certainly not excluded, (ii.) Three forms, 
including (a) 0 . biennis in the general sense, (b) 0 . mnricata, and (c) 
a large-flowered form belonging to 0.grandiflora or 0. Lamarckiana, 
were clearly recognized and figured by Barrelier in 1714; (c.) was 
probably the same as the large-flowered form of Ray. (iii.) The 
early synonomy gives clear evidence that this large-flowered form 
was cultivated in many Botanical Gardens in pre-Linnaean times, 
(iv.) An CEnothera which belongs to 0. Lamarckiana in the rather 
narrow sense is now the commonest form in English gardens, where 
it seeds itself from year to year, (v.) A very similar form has been 
found by Heribert-Nilsson (24) in the gardens of Southern Sweden, 
(vi.) A form which is practically identical with that of De Vries has 
been flourishing in abundance in a naturalized condition on the 
Lancashire coast since at least 1805. 
These facts equally permit two explanations. (a) That 0. 
Lamarckiana as it now exists in England, naturalized and in 
gardens, is essentially the same plant described by Ray and by 
Barrelier and is descended directly from seeds brought from 
“ Virginia.” We now know it to be probable that such seeds would 
not produce a uniform race to begin with, (b) The other possibility 
is that the race above-mentioned from Virginia belonged to O. 
grandiflora rather than to 0. Lamarckiana, the latter originating 
through crossing afterwards. One weakness of this theory is the 
fact that O. biennis races also occupy this area of Virginia which was 
