OBSERVATIONS ON INDIAN PRIMULAS. 



319 



been collected previously, but confused with P. obtusifolia, var. Griffithii. 

 It is a good species found in Rhododendron glades and possessed of very 

 beautiful pale lavender-coloured flowers. 



P. Kingii, Watt, is a lovely plant met with in Sikkim. It has 

 leaves shaped as in some forms of P. purpurea, only smaller, thicker, and 

 almost leathery in texture. Flowers usually pendent, and of such a dark 

 claret colour that they are almost black. This species, I believe, would be 

 much admired were it introduced into cultivation. I found it in full 

 flower at 14,000 feet, and most collectors have done the same, the result 

 being that we have not as yet got the seed. It has been collected in Tibet 

 by Hobson. P. Gammieana, King, I believe to be at most a variety of this 

 species, and still another allied form, P. ametlujstina, Franch., has come 

 from Yunnan. 



P. Elwesiana, King, is perhaps the most striking Primula of Sikkim. 

 It occurs at altitudes of 12,000 feet. It has large, solitary, deflexed flowers, 

 borne on much-elongated, thickened, and pilose peduncles, destitute of bracts. 

 It is the representative of a Chinese group of great beauty, of which P. 

 vincceflora and Delavayi of Franchet are superb examples. These have 

 recently been procured from the mountains of Yunnan, and may yet with 

 further study be found to constitute a subgenus. They recall in some 

 respects Bryocarpum. 



Perhaps the best-known cultivated example of a Primrose belonging to 

 this section would be P. nivalis, Pall., especially the var. turkestanica. It 

 is found in Turkestan and Persia, and thus keeps up the character of this 

 assemblage, being strongly N.W. Himalayan. The form of P. nivalis 

 ■collected at St. Matthew Island during the British Behring Sea 

 Commission still further preserves this peculiarity. It might be 

 ■described as closely allied to variety Moore roftiana. And, as if to confirm 

 the reduction of P. Stuart ii to P. purpurea, there is a yellow-flowered 

 form of P. nivalis that has been described by Regel as var. Bayemi. 

 The Altai form of P. nivalis corresponds closely with the leathery-leaved 

 form of the Indian condition, for which at one time I proposed the 

 name P. plantacjinca — a smaller plant, with narrower leaves than 

 P. Moorcroftiana. 



5. Petiolaris (fig. 73). — This in more senses than one may be described 

 as the most sportive assemblage of Indian Primulas. The species thrown 

 together under it are not only found to vary freely, according to soil, 

 ■exposure, altitude, &c, in which met with, but they obey the dictates of 

 ■cultivation almost instantly. The central feature that separates the 

 group may be said to be the presence of a distinct petiole in place of the 

 spathulate-cuneate sheathing base of the leaf met with in the majority of 

 the species placed in the other sections of this classification. The name 

 petiolaris at once suggests that peculiarity, but, strangely enough, it has 

 been given to the species of the assemblage that is least petiolate, namely 

 P. petiolaris. There have been described in the " Flora of British India " 

 seven varieties of that species, but with a very little stretch of imagina- 

 tion that number might easily be doubled. In three of these varieties the 

 leaves are usually obovate-spathulate-sessile, but occasionally a rotund leaf 

 borne on a long naked petiole may be found. In the other varieties, 

 petiolate leaves are universally present, along with spathulate sessile 



Y 2 



