REPORT OF WORK IN 1915 IN KANSU AND TIBET, 329 
Adonis coerulea (F 518) also yielded only very few and imperfect 
seeds. It is too leafy a little plant to be admirable, and is common 
all over the lower alps of the region. Its blossoms, being of a satiny 
mauve or violet, long puzzled me as to its name, until I discovered 
that in all its forms coerulea does indeed dry to a lovely tone of 
turquoise blue. 
Androsace tibetica (F 533) is, of course, the same as last year's 
F 246, but I think better to give it a fresh number, as the regional forms 
of this species are so many. That of the Da-Tung represents two main 
lines ; up at Wolvesden House it is squat and broad-leaved, with 
flowers especially large and fat and round ; down at Tien Tang, on 
the hot gravelly banks, and in the parched lawns, it is of so intense a 
rosy crimson and rich pink that I wonder yet again to think how ever 
I was led to describe the plant originally as white, before I came to 
know it. A further fact emerges also from the imperfect information 
of last year, and from subsequent experience with seedlings. And 
that is that A. tibetica and A. longifolia cannot easily be grown in soil 
too poor or in positions too hot and dry. Cosseted in nice fat, well- 
watered pans at least, they both of them grew very yellow and sickly 
with me, and when A. tibetica flowered it was with a pallid ansemic 
bloom in no proportion to its lush unhealthy leafage ; while A . longifolia, 
its yet more torrid cousin, was obviously dying altogether. And they 
both only began to pick up when got out under a hot wall, and guarded 
strictly from the watering-can. Without such treatment A. tibetica 
will, I think, usually prove pale and hypertrophied, and A . longifolia, 
by far the more difficult species, evidently, of the two (demanding 
much hotter and harder conditions), will continue as impossible of 
culture as I once began to fear it. It should be fully remembered 
that these plants are not to be venerated as typical children of the 
alps, like their cousinhood, but regarded as the race's special kindliness 
for the needs of hot south-country English gardens. A. tibetica Mariae 
I was not able to get in seed, and for a long time the identity and 
whereabouts of this narrow-leaved form were a puzzle to me. But 
finally Purdom found that it filled the marish flats about the Koko-nor 
with sheets of rosy blossom, long after the broad-leaved type had 
grown fat in seed at Wolvesden. (Photograph and painting secured.) 
A. mucronifolia (F 319) will ere-long be distributed in propagated 
plants. This is one of the few triumphs of our packing, and I luxuriate 
in pleasure each time I see my two or three specimens in pots, thriving 
as heartily as if they had no more memory of their own far alps, and 
remember what poor straggling messes they were when, at the end of 
my first season, they were despatched from Lanchow in the winter 
of 1914, across the changes and chances of distance and posts and 
railways. A. mucronifolia abounds in the Da-Tung Alps, but lower 
and more especially on their rare limestone outcrops. It is not here, 
however, quite the same be-snowed entrancing loveliness that it is 
in the highest fine lawns of green on the topmost necks of the Min 
S'an. (Photograph.) 
