336 JOURNAL OF THE ROYAL HORTICULTURAL SOCIETY. 
I returned to Tien Tang, so that its quantity leaves much to be desired, 
and possibly many desiring ; as for its prospects in cultivation, one 
can but hope. 
Incarvillea sp. (F 509) is almost certainly /. grandiflora, and the 
same as F 89 and F 268 (though this cannot be taken yet for sure). 
Anyhow this is a very handsome species, from rocks and open banks 
up to ten thousand feet in the Da-Tung Alps, scattered through the 
range, but not so abundant as the earlier numbers last year. As seen, 
it was always monocarpic, and a finer, more exotic, gaudy thing could 
hardly be imagined. (Painting and photograph.) 
Isopyrum sp. 504 should by all obvious rights be /. gvandiflorum . 
But in that case what becomes of last year's plant, also presumably 
I. gvandiflorum ? For, while that was a pretty pallid thing enough, 
this, with Gentiana sino-ornata, stands out not only as the loveliest 
thing in the year, but almost in all the years of my . collecting any- 
where. In the granitic, volcanic,*or calcareous cliffs of the Da-Tung 
range, from eleven thousand feet upwards, it forms great tuffets of 
fine glaucous foliage, from which float in June the most glorious big 
flowers of richest lilac purple, in size and colour suggesting a 
glorified Anemone nemorosa Allenii, but of inimitable silken texture. 
It is invariably restricted only to cool shady aspects and cold shaded 
faces of the cliffs ; and only at its topmost limits is found out on 
the open boulders, though even there for preference on the cooler 
faces, and dwarfed and compacted with the elevation. And it is as 
rigidly saxatile in habits as Phyteuma comosum. I was fortunate in 
getting an abundance of seed (indeed, it was the " clou " of this year's 
collecting), as such a plant will be none too easy to raise, I suspect, 
and likely to prove pernickety even afterwards as to its position and 
the processes of getting it there. (Photograph and painting.) 
Lancea tibetica was sent under two numbers, as F 541 and F 670. 
In neither case was the seed satisfactory, having, by the necessities 
of our movements, to be collected in pods still immature. It is the 
little thing I talked of in 1914 as Mazus sp., and abounds all up the 
March, evidently, for choice, in bare open banks and flats at mid- 
alpine elevations, where it ramifies into lax carpets of dark-green 
rosettes, in the midst of which nestle in July the rich violet-purple 
helmets of the blossom, giving place to fat pods that in time blush 
to a deep varnished crimson, as they very slowly come towards their 
ripening. 
Leontopodium. — The common Flannel-flower of the European 
Alps has its absolute counterpart in the high lawns of the Da-Tung. 
But I only sent two members of the race. Of F 741 I doubt whether 
it should rightly be Leontopodium or Antennaria. All over the loess 
region, from Wei-yuan Pu throughout the foothills of the alps, and 
up to ten thousand feet, it everywhere forms, in the short-cropped 
stretches of grass, wide, perfectly flat, and tight scabs of silver grey, 
with innumerable rosettes of foliage, from which, on stems of barely 
an inch, unfold a galaxy of small Edelweiss stars in June, and on 
