369 
essential character, though the specimens are mostly stouter, with larger 
bulbs and a shorter suprabulbous axis. Their leaves are also more 
obtuse pos those of the Kansu specimens, but one of these exhibits 
exactly the same foliage. 
Przewalski and Potanin's specimens were collected “in pra 
alpinis humidis rupibusque totius Kansu Delavay describes the 
habitat of the Yun-nan violet as “in pascius pinguibus et humidis," and 
Lace indieates the plants as growing **in short turf amongst the gau *. 
The elevations at which Viola bulbosa was found are: about 10,500 
feet in Bussahir, 12,000 feet in Sikkim, 8,800 (?) feet in serai 
10,000 feet in Yun-nan, and 9,000-10,000 feet in Kansu 
The peculiar mode of .v vegetative reproduction by bulbs and their 
structure, have been very accurately described by mowiez. The 
bulbs, which sometimes attain the size st 4 a pea, aye of a short 
and fleshy axial portion, and of 4-8 scales, which are very fleshy at the 
base, but scarious near the margin. The scales are rotundate, obtuse, 
or shortly Wir eR and the upper ones sometimes have short scarious 
stipules attac to them. They are etr the modified basal 
portions of mie much reduced leaves. Abov e bulb the axis con. 
tinues, forming a short or more or less elo ated and slender stem 
The bulbseales, as well as some of the leaves following next to them, 
produce from their axils buds which are more or less flattened from the 
back and ovoid. Maximowicz calls them “ bulbilli," although they 
hardly m this name. ‘They grow into slender stolons, which, apart 
rom a few minute scales, are naked, and quite like the young sto tolons of 
other violets, as for instance, Viola uliginosa. ey are 1-2 in. long in 
the Kansu and Sikkim specimens, In the Bussahir and in the Yun-nan 
specimens they are apparently absent, but they may be found on a 
closer examination as more or less developed buds, laiia 3» the 
bulbseales or the leaf sheaths. Maximowicz su ggests that these 
stolons become finally transformed at their apices into tre bulbs which 
emit a leaf and flower-bearing axis in the next season. Although there 
are no specimens exhibiting these bulbs in their stage of res t in the 
Kew Herbarium, the plants collected by Lace are very much in favour 
of this suggestion, as a portion of what I take to be a stolon is still 
attached to the bulbs in several cases. The rootiets necessary jd the 
independent establishment of the young plant spring in great number 
from the base of the bulb and the axils of the bulbscales, and if the 
as far as my g. et quite unique in the 
Viola bulbosa was compared by Maximowicz with "Vila Patrinii, 
.C., and, indeed, it seems to approach closer to it than to any other 
species of Central Asia or the Him malaya, in spite of its very peculiar 
innovation. True, the stigma of Viola bulbosa is almost e exactly as in 
i 
being flattened at the back. The stigma itself has two very small 
rounded lateral wings, which, in combination with a slight depression in 
front, give it the peculiar shape of an easy chair with a very low back. 
Below this depression and above a very minute triangular hyaline lip 
