Hybridism and Classification in the Genus Rosa. 155 
generally agreed in holding R. glauca Vill. to be as good a “ species ” 
as R. lutetiana or R. dumetorum. A knowledge of the anatomical 
construction of plants is useful and often very helpful, hut taken 
alone it is not always sufficient to solve the problems of relationship. 
One of the most striking results of the intensive study of 
British plants is the increasing number of hybrids that are being 
recognised. One need only mention such genera as Viola, Epilobium, 
Mentha and Salix. The presence of hybrids renders the classifica¬ 
tion of such genera extremely difficult, and it is not improbable that 
the difficulty in classifying roses may be largely due to hybridi¬ 
sation and segregation, complicated, it may be, by re-hybridisation. 
When one considers the general prevalence of wild roses, the 
frequent association of numerous species and forms in the same 
habitat, the large numbers of flowers produced by fully developed 
plants, it becomes evident that there must be abundant opportunities 
for natural crossing. 
Several hybrids, in the sense of crosses between distinct species, 
have been recognised among British Roses. The commonest is 
R. involuta described by Smith, FI. Brit. 1804, from material 
collected in the West of Scotland. It was not until 1884, however, 
that its hybrid origin was suggested, when Christ promulgated the 
ideathati?. involuta Sm. is a natural cross between R. pimpinellifolia 
and R. tomentosa. This view is now universally accepted. R . 
involuta , though not common, is fairly widely distributed in the 
North of England and in Scotland, and it may be expected to occur 
wherever the two parents grow together. Numerous forms of the 
hybrid are to be found and many of these had received varietal 
names before the hybrid nature of this rose was realised. While 
R. pimpinellifolia shows little or no variation, R. tomentosa is 
decidedly a polymorphic species and considerable variation in 
R . involuta need not, therefore, be unexpected. In general, the 
characters of the hybrids are more or less intermediate between 
the unlike characters of the two parents, but some of the forms 
bear a greater resemblance to one or to the other parent, and it is 
not impossible that this may be due to R. pimpinellifolia being the 
seed-parent sometimes and R. tomentosa at other times. The hybrid 
plants are usually vigorous, flower freely and are more or less sterile, 
thus showing some of the attributes of hybrids between distinct 
species. Whether these hybrid forms of R. involuta breed true or 
whether they segregate is unknown, but the work of Brainerd 
(1906) on Viola suggests the need for similar investigation in the 
case of our natural rose hybrids. 
