EAST NEPAL AND THE SIKKIM HIMALAYA MOUNTAINS. 73 



(serpens and others). These are all provided with soft woody 

 swellings on the root, of all sizes, from that of a nut to the thick- 

 ness of a man's thigh, which, though structurallj^ the same as 

 other parts of the root, serve as reservoirs for a great quantity of 

 fluid destined to nourish the plant in the drier season of the year. 

 These plants never grow on the ground, properly speaking, but 

 often on exposed rocks, where the use of these tubers is more evi- 

 dent, and it is a part of their economy to be so provided. The 

 Rhododendrons have never such organs, and there is no difference 

 between the root of a specimen grown on the ground and one from 

 a mossy tree-trunk. Such species as are habitually epiphytical re- 

 quire a lighter soil, with plenty of moss, and a very damp, humid, 

 equable temperature ; and will, I am sure, present no insuperable 

 obstacles to the cultivator. 



Soil. — There is in this respect little variety throughout Sikkim, 

 and, as far as vegetation is concerned, it may be divided into vege- 

 table mould and stiff clay — each, as they usually occur, remark- 

 ably characteristic in composition of such soils. 



The clay is uniformly of great tenacity, and is, I believe, 

 wholly due to the effect of the atmosphere on crumbling gneiss 

 and other rocks. It is tenacious, seldom friable, and sometimes 

 accumulated in beds 14 feet thick, although more generally of 

 only about 2 feet. In certain localities, beds or narrow seams of 

 purer felspathic clay on vegetable matter occur in it, probably 

 wholly due to local causes. An analysis of that near Darjiling 

 gives about 30 percent, of alumina, the rest silica, and a fraction 

 of oxide of iron. Lime is wholly unknown as a constituent of the 

 soil, and only occasionally seen as a stalactitic deposit from a few 

 springs. 



The vegetation has always good drainage, from the broken 

 nature of the subjacent highly-inclined stratified rocks : with 

 the Rhododendrons of the second zone this is especially the 

 case, and they thrive luxuriantly on the soil overlying old 

 moraines. 



A layer of vegetable earth almost invariably covers the clay 

 for the depth of 3 to 12 or 14 inches. It is a very rich black 

 mould, held in its position on the slopes of the hills by the 

 dense vegetation, and accumulated by the banks of small streams 

 to a depth at times of 3 and 4 feet. R. arhoreum is one of 

 the few species that seem to avoid this soil, but, I think, in 

 appearance only — that species loving exposure, and therefore 

 occupjang a comparatively poor dry soil, over which other things 

 do not extend. I have seen R. arboreitm and barbatiim growing 

 on the margins of pools of water, in what is almost bog earth ; 

 but pools and bogs are rare in Sikkim, and of very local occur- 

 rence. The following is an analysis of an average specimen of 



