APPENDIX. 



was in reality a distinct species of similar stature, but with densely white- 

 powdered foliage (P. dealhata, 1915, Sp. nova) ; which leaves me suspecting 

 that the same may ultimately have to be said of another so-called " form " 

 from Monk Mountain. Accordingly, I have labelled all sendings of P. 

 " stenocalyx " with the name of their district ; and advise that all these be kept 

 apart and carefully watched, as possibilities are so many. It even, I believe, 

 will be found to include a few stray seeds of P. No. 20, from collected crowns 

 sent down with the true cognata's, to ripen their pods in J6-ni. In the earliest 

 lots, however, which alone were large enough for general distribution, I am 

 certain that P. stenocalyx will be found pure, and possibly unalloyed except for 

 the Lotus Mountain plant, which undoubtedly comprises the majority, if not 

 the whole, of the second sending. (P. cognata is a chimaera, it seems.) 



Primula Sp. No. 20 (F 196) is blurred with the last, and very scanty in 

 supply, even if sent at all. It need not be regretted ; it is a starveling little 

 thing, replacing No. 2 in the highest cool cliffs and grassy rock-ledges of the 

 uppermost Min S'an. It has the puny look of P. yunnanensis — a feeble tiny 

 rosette, and a scape of an inch, more or less, with 2 or more flowers. These 

 we never saw, unless some rather attractive starry recurving blooms of lilac- 

 mauve from the great Ardjeri gorge did indeed represent this species in a stout 

 and drawn-up form (for here the scape had attained 3 or 4 inches, and 

 the abundant crowns seemed stronger than up above. It was here growing in 

 damp cool silt, very loose, about the feet of great boulders in the shade, at the 

 mouth of the ravine). 



Primula Sp. No. 21 (F 197) may perhaps contain two species, of which only 

 seed from Rou Ba Temple has been distributed. As I know the plant, in the 

 cool silty grottoes and shady boulders of the Ardjeri gorges, it stands in very 

 close alliance to P. lichiangensis, precisely repeating its habit and foliage, and, 

 though I think it distinct, differing distinctly for the worse, in rather anaemic 

 pallid stars of blossom — though this may be only because the flowers were 

 then, at the beginning of August, in their final stage of disappearance. In 

 any case, P. lichiangensis gives the picture and the rule for this and also 

 (whether it be the same or no) for the parent of the other seed sent under this 

 number — a woodland species from forest banks about Rou Ba Temple and 

 opposite J6-ni. It is distinct from P. Silvia, and a quicker, freer grower, 

 though inferior. 



Primula Sp. No. 22 (F 248). — For the differences between this superb species 

 and its smaller cousin, see under P. No. 10. P. No. 22 makes robust and clod- 

 forming clumps of stiff upstanding foliage all over the gaunt consolidated silt 

 beds and hard earthy shingles of the uppermost aretes of the Min S'an, in the 

 same sort of places chosen by P. No. 10 on Thundercrown, but growing much 

 stouter and more abundant, often making quite a waving jungle of its stalwart 

 stems over gaunt slopes where no other living thing occurs. It weeds up in 

 sods like a groundsel, and roots in the same rampageous manner as P. No. 10, 

 with the same long, pale, and chaffy pods, though I fancy it more rarely super- 

 imposes a second flower- tier on the first. The flower is so far unknown ; judging 



519 



