. A NEW BRITISH SAXIFRAGE. 569 
Account of a new Brivish Saxtrrace; by W. H. Harvey, M.D. 
&c., Professor of Botany to the Royal Dublin Society. 
(With a Plate, Tas, XIX.) 
The announcement of a new British Saxifrage carries, on the 
face of it, a mark of doubt; especially as the one I have to intro- 
duce belongs to the group of 8. umbrosa, a group almost pro- 
verbially variable and uncertain in a variable and uncertain genus. 
I must also admit that our new plant was not originally found in 
a flowering state, and has produced the only flowers which have - 
been seen after having been cultivated for three years in a garden. 
This circumstance, for the present, may prejudice many persons 
against receiving the new plant into the calendar; but if not a 
good species, a point which I leave to botanists to decide, all 
must admit that it is at least a very remarkable variety, and as 
such, is worthy of being figured, and of having attention directed 
to it. The shape of the leaves is very peculiar. They are much 
longer and more spathulate than those of any other of the umbrosa — 
group that I have seen, and almost remind one of those of 
S. cotyledon and its allies. But distinctions derived from the 
leaves are not those on which, in this genus, I am disposed to 
place much reliance, for it must be owned that the leaves of 
S. umbrosa, 8. Geum, and their allies vary extremely in outline ; in 
the length of the petiole, in the crenatures of the margin, in pubes- 
cence, in short, in all their characters. This new species (or . 
variety), however, is chiefly characterized by differences in the 
structure of the flower, and these are so marked, that it can 
scarcely be placed in the same section of the genus as S. umbrosa, 
but rather belongs to the group of S. nivalis. In the wmbrosa 
group the calyx is parted to the base, the sepals are perfectly free 
from the ovary, and are strongly reflexed soon after the expansion 
ofthe flower. In our new species the calyx is gamosepalous, | 
cleft two thirds of its length, the tubercular portion adheres to 
the base of the ovary, and the limb, instead of being reflexed, is 
simply spreading. Add to this, that the petals are much broader 
and more elliptical than in any of the group, and are elegantly 
dotted over the whole surface, and we have characters sufficient, 
