255 
necessary in not too hastily concluding, when anthesis takes place 
simultaneously, that we have more than one species, however 
diverse the forms may appear. 
The mature fruit has not received the attention it deserves; 
likewise the seed. I am satisfied that from them both important 
aids can be derived in the differentiation of species. Inthe early 
part of winter the hips of Rosa Carolina wither and wrinkle, 
forming a short but distinct neck; the remaining species of the 
group Caroline, on the other hand, so far as my observation 
goes, continue plump until the following spring. By this char- 
acter alone, and especially when taken conjointly with the seed, 
which is smaller than in either of the other species, Rosa Carolina 
may be known even when divested of flowers and leaves. In like 
manner, as the result of close study, the fruit and seeds may 
afford the means of distinguishing other species and putting them 
on a more permanent basis. 
These remarks are intended to be simply suggestive and the 
end attained for which they are made if they stimulate more 
earnest work on the part of botanists in solving some of the 
problems connected with the genus Rosa. In the same spirit 
I have thought proper to advise the dropping of Rosa lucida, 
Ehrh., as a species, and placing it under R. humilis, Marshall, as 
a variety. They commence flowering at the same time; inter- 
mediate forms occur that connect them so closely that among a 
large number of specimens it is well nigh impossible to know to 
which species some should be assigned. They differ only in a 
few variable characters which are far from being constant. It is 
therefore presumed that the best interests of the science are sub- 
served by putting them together. By this arrangement all those 
forms which show a transition to Rosa Carolina, L., and yet in 
reality are distinct from it, are placed under R. humilis, to which 
they are more closely related. 
There is another form found growing on the margins of damp 
woods, or in the clearings, which in some respects seems inter- 
mediate between Rosa humilisand R. nitida, Willd. This I have 
also placed under the former as var. vz//osa. It remains for future 
observation to assign its limits; it is not infrequent in New 
Jersey. 
