EAST NEPAL AND THE SIKKIM HIMALAYA MOUNTAINS. 75 



mould of loam and peat, with good drainage, are necessary for 

 the shrubby plants of these regions. 



On the Species. — Since the preparation of the several parts of 

 the illustrated work on the Sikkim Rhododendrons, my opinions, 

 which were then far from fixed as to the limits of some of the 

 species, have changed, with the many additions to the localities, 

 &c., I have collected. When the first part was publislied, I had 

 only seen single specimens of some of the species figured ; and 

 without books of reference of any kind, I was unable to deter- 

 mine their names ; the specimens and drawings sent diflTered 

 so widely from cultivated individuals, and from the indifferent 

 herbarium specimens attainable, that they were pronounced new, 

 and published accordingly. The results of a careful study of all 

 my species and specimens are, that of those figured in the first 

 parf, R. Campbellice does not differ from R. aThoreum^ R. lanci- 

 f'olium from R. barbatum, R. Wallichii from R, campanulatum, 

 and R. Roylei from R. cinnabar inutn. 



In the second part a synoptical view of all the Indian species 

 is given, in which some of these errors are indicated, and the 

 species are thrown into groups defined by characters I still think 

 natural and constant. In this and in the concluding (third) part 

 22 more Rhododendrons are figured as species, several of which 

 were regarded by myself as well-marked forms or varieties only, 

 and as such worth figuring, but which were published with the 

 provisional specific names I attached to them. To these I shall 

 allude liereafter. Lastly, I have gone over the Bhotan collections 

 of Mr. Griffith, and identified all his 14 species with my own, 

 except the one figured by Dr. Wight as R. grande. Of this I 

 find no specimen in the set of Griffith's plants given to Sir W. 

 Hooker, except it be the R. Hodgsoni, which I can hardly re- 

 concile with the plate and description in Dr. Wight's Icones. 

 Dr. Wight's R. Griffithii, on the other hand, 1 find to be 

 founded on a very starved state of R. Aucklandii, a name which 

 I waive. 



Bhotan borders upon Sikkim to the eastward, and presents the 

 same characters of humidity in that western part (wiience Mr. 

 Griffith's specimens wei'e procured) as Sikkim does. Its lower 

 and outer ranges of liills, however, being drier (from reasons 

 hereafter to be explained), it does not appear to be so rich in 

 Rhododendrons, and hence probably no peculiar species are 

 found; for to the eastward of Bhotan again, in the " Durrung" 

 country, other species have been discovered by Mr. Bootli. 



I shall now give a list of tlie 27 Sikkim species known to me, 

 arranged in natural groups, with such brief characters as are ne- 

 cessary for determining them when they shall flower. An account 

 of the geographical features of Sikkim will follow, and a division 



