Transactions. 107 
very extensive and carefully botanised area. It is perhaps too 
soon to speak definitely of the distribution of the roses in our 
district, but apparently the greater number are met with in the 
main valley of the Nith, and often many varieties I have observed 
grow together in a small area. They appear to thin out some- 
what in the side valleys, and the fewer number reach the base. of 
our higher hills. Although from this height some are found with 
Rosa mollis, notably subcristata and dumalis descending through- 
out the range to sea level. The study of roses is rendered more 
difficult and confusing from the occurrence of approximate and 
intermediate forms. I have to acknowledge invaluable help from 
Mr Bennett of Croydon, in naming specimens and for his kind- 
ness in sending a collection to Mr Baker, of Kew, for inspection. 
T append a full list of the genus as found in Nithsdale by Dr 
Davidson and myself, which will bring it up to date and facilitate 
further search. In it are included a few intermediate forms 
which may be found interesting, having had the best authority 
for their recognition. \ 
LIST OF ROSA. 
1. Rosa spinosissima, Linn. 
. mollis, Sm. 
bo 
3. ,, var. cerulea, Woods. 
4. ,, var. pseudo-rubiginosa, Le}. 
5. ,, tomentosa, Sm. 
(3). alae hy var. swbglobosa, Sm. 
{caus . var. scabriuscula. 
8. ,,  rubiginosa (an escape 4) 
9, ,, canina, var. lutetiana, Leman. 
10. ,, M ,, duwmalis, Becist. 
(ayes, e H : f. Malmundariensis, Le}. 
ae 7%, Af » wurbica, Leman. 
JSP anes a: F 4 f. platyphylla, Rau. 
Vay,, 4 » arvatica, Baker. 
HOU! i. 5, dumetorum, Thuill 
AGy!',; , x 5 f. wneinella, Besser. 
1 ae i » pruinosa, Baker. 
Si; 1 5, encana, Woods. 
WOST 2 » 97 bomentella, Leman. 
20P™,, a », andevagensis, Bast. 
YA aoe MY 5 verticillacantha, Merat. 
