REPORT OF WORK IN 1915 IN KANSU AND TIBET. 345 
the advance seeds in the Da-Tung. This he found in a marish grassj' 
hollow on Kweite Pass, and I have nothing to say of it but that this 
also does not quite look as if it belonged either to P. tangutica or 
P. Woodwardii. Neither of this, of course, nor of its predecessor, was 
any trace of flower remaining when the plant was found. Otherwise 
the tale would be plainer. 
Groups of Primula represented : 
Auriculata — farinosa. P. sibirica (?), P. stenocalyx, P. gemmifera, 
P. Reginella. 
Sinensis. P. sinensis. 
Nivalis — Maximowiczii. P. Woodwardii, P. tangutica, P. Farreri. 
Souliei. P. urtici folia. 
Rosa sp. (F 544) is, I think, the best of the year's shrubs. I only 
saw it at one point, in the shingles of the Da-Tung Hor, where it makes 
its great sweep round from Bridge-head towards Tien Tang. Here, 
in the hot stretches and gravelly steeps, the Rose made fine elegant 
bushes of five feet or so, slender and graceful, with small-folioled 
greying foliage, and bloomy young shoots of pink. The flowers are 
very profuse, very fragrant, and of clear rich rose, all along the sprays 
and arching boughs in small clusters, followed by no less brilliant a 
show of glossy vermilion berries, bead-shaped, and shedding their 
calyces so quickly that they look more like some Cotoneaster's. It 
is a shrub of quite particular charm, obviously in the alliance of 
R. Webbiana and R. Willmottiae, but, as it so far seems, distinct 
from both. 
Rosa sp. (F 774) is probably that little fine many-flowered rose 
with long, narrow haws, sent in 191 4 as F 84, not uncommon in the 
lighter open scrub all down the March of Kansu. 
R. sp. (F 755) , on the contrary, does not begin to appear till you 
are down over the border in Northern Szechwan, where, in company 
with that pervasive glory of all those regions Rosa sp. F 291 (only 
met with in 1914 in the farthest western extremity of its distribution), 
it occurs much more rarely amid the shrubbery on the long high 
hill-tops of the Red Basin. This also is a big rampant rose, with high- 
arching boughs. The flower is unknown, but red fruits are borne 
in loose clusters along the boughs, and the infrequent leaves are 
large-folioled and very handsome, dark leathery green, and clothed 
on their reverse with a dense soft velvet of pubescence. (I know it 
as the Velvet Rose, accordingly.) 
Rosa sp. (F 783) is that lovely Golden Rose which I failed to get seed 
of in 1914, but which last year Purdom captured on his way down 
through the Western March to meet me on the Szechwan frontier. 
The round dark fruit appears to fall untimely, a trick I had not sus- 
pected, but had attributed its disappearance to birds. Its shape and 
colour lead me to believe that this conjectural species is after all 
R. xanthina (though I see Mr. Bean is inclined to deny that R. 
xanthina has any high claim to be recognized as existing at all). On 
the other hand, it is assuredly not the much more pallid R. Hugonis 
that I had thought it at one time. And so I leave it ; those bending 
