196 JOURNAL OF THE ROYAL HORTICULTURAL SOCIETY, 



OBSERVATIONS ON INDIAN PRIMULAS.* 

 By Sir George Watt, CLE., M.B., CM., LL.D. 



[Read before the Horticultural Club, October 4, 1904.] 



Within the past few years remarkable progress has been made, both by 

 botanists and cultivators, in the discovery and production of new 

 forms of Primula. There are now known to exist close on two hundred 

 species, f besides many cultivated hybrids and sports of great beauty. 

 Roughly speaking, they may be said to be distributed within three 

 great centres : namely, Europe, India, and China. Each of these 

 chief centres possesses on an average about fifty species, and the 

 balance of the total mentioned may be taken as made up by America, 

 Central Asia, Africa, and Japan. They inhabit the temperate and 

 arctic regions of the northern hemisphere, practically only one species 

 — P. magellanica — -being found in the southern hemisphere. This 

 circumstance is perhaps the most remarkable feature manifested by 

 the genus, but there are others equally impressive. P. magellanica, 

 for example, has two great centres (New Mexico and Patagonia), 

 separated by five thousand miles in which no trace of that plant can 

 be discovered. But P. farinosa has the widest distribution of all 

 the species, since it practically occurs in every important Primula area 

 throughout the northern hemisphere, and yet in its diversified and 

 wide distribution it manifests but the very slightest modification 

 either in form or in size. Of the Indian species all without exception 

 are confined to the Himalaya and adjacent mountains that form 

 the North-Eastern, Northern, and North-Western frontier of India, 

 from Burma and Assam to Kashmir and Baluchistan. None occur 

 on the mountains of Central and Southern India, though in point of 

 climate, soil, and associated plants they might fairly well have been 

 looked for in these regions. So again a few species are practically 

 met with all along the mountainous frontier of India, while others 

 are exceedingly local. A few forms are very variable ; others seem, 

 like P. farinosa, remarkably constant. But Sikkim may without 

 hesitation be pronounced the headquarters of the Indian Primroses, 

 and many species are found in that country that would appear to 

 occur nowhere else. The forms that spread eastward from Sikkim 

 to Assam, Burma, and Manipur are seen to belong to an assemblage 

 that attains its greatest development in Southern China, more especially 

 in the mountains of the province of Yunnan. On the other hand, 

 the types that spread westward from Sikkim to Nepal, Kumaon, 



* Reprinted from Journal R.H.S. xxix. (1904), p. 295. Several plates 

 illustrated this article on its first publication which are not reproduced here, 

 f Cf. the more recent estimate, p. 129. 



