APPENDIX. 



with Primula and Pleione, we all adventured back again over the hills to 

 Gahoba and Satanee, pleasantly and placidly exploring the Satanee Alps till 

 at last the alpine winter came crashing finally down on 18th October, and the 

 snowfall drove us back again to sunny, torrid Siku — the change being equivalent 

 to that of Torbole or Garda from the Stelvio. From Siku we made our final 

 ascent of Thundercrown in quest of Delphinium tanguticum, and then Purdom 

 hurried north again for odd jobs in the Tibetan Alps, while I remained to finish 

 a few final collections at Siku. On 30th October I also bade a last sad farewell 

 to this delicious little corner of sunshine, and rode northward to join Purdom 

 at a small village nearly opposite J6-ni, whence on 13th November, the country 

 and the green hills being now all gone brown and sere and dead in winter, we 

 all moved up northward over Lotus Mountain towards Lanchow, there to spend 

 the dead months, after a well-rounded season, in which, after all its storms 

 and stresses, there have been only two downright failures to regret — Iris Henryi 

 and Farreria Sp. — neither of which was it possible to attempt. 



As for such cultural hints as I give, these, of course, are purely conjectural, 

 and based on my local observations. The foregoing itinerary is meant to sug- 

 gest the various climates of the districts I have this year explored, and the 

 stations given for various plants will serve to identify each with its own con- 

 ditions. Generally speaking, these northerly ranges should give no such legacy 

 of tenderness as is bequeathed by the warm wet atmospheres of Yunnan and 

 Szechuan that have bred us so many disappointments. The Satanee range has 

 a climate close akin to our own north country conditions, with a very hard 

 winter. Yet harder is the winter and damper the conditions in the vast grass- 

 lands of the Min S'an Alps. Between the two lies the hot loess region of Siku, 

 and it is from there alone that we may expect its plants to want favouring 

 circumstances of drought, heat, sunshine, and a hard stony soil. The Thunder- 

 crown ridge, however, with its daily shower, stands far above the circumstances 

 of the hot loess at its feet ; and its children will take the culture of the general 

 high-alpine flora of the Min S'an, to which indeed they belong, though cut off 

 from their kindred upon a remote and insulated mass of limestone. And this 

 last word reminds me of a last caution. For whatever the information may be 

 worth, both the Min S'an and the Satanee range are essentially calcareous, so 

 that, except where a special caution is entered, it may be taken that all the 

 following plants are calcicole in nature. 



Adenophora.— These beautiful Campanulads take the place of their august 

 cousins in the alpine grass-lands of Tibet, having all, more or less, the habit of 

 Campanula rhomboidaUs, though with longer and stirrer displays of bells. One 

 common species has showers of poor little almost globular flowers, with far- 

 protruding style ; ;i second is an improvement, leading on to the next, F 235 

 (? A. Potaninii), a really lovely thing which should be of the easiest culture in 

 any healthy open place. It has crisped, deeply dentate, incanescent foliage, 

 and noble branching panicles of big blue bells, far better furnished and more 



490 



