Prof. Don's Descriptions of Indian Gentianeae. 505 



Mem/anthes and J'illarsia, lias been recently separated from Gentianece on 

 account of their alternate lobed or crenated leaves, characters which appear 

 to arise from the peculiar circumstances under which tiie plants live, and 

 perhaps of as little importance as the entire absence of those organs in the 

 parasitical genera f'o/tiria and Leiphaimos ; and the arrangement of the leaves 

 is of less importance, since they are alternate in two species of Swertia. 



I had formerly proposed (Edinb. Phil. Journ., July 1831, p. 275,) to refer 

 the remarkable genus Desfontainia* to the Gentianece, but from the circum- 

 stance of its possessing a multilocular ovarium, deciduous corolla, with im- 

 bricate aestivation, undivided stigma, opposite, spinously toothed, penninerved 

 leaves, it is evident that the view which I then took of its affinities was erro- 

 neous ; and I think it not improbable that it will be found to be more nearly 

 related to Ericaceie than to any other family. In my description I have de- 

 scribed the berry as unilocular, with 4 or 5 parietal placentae, but I now find 

 that it has the cells comjiiete, and is therefore multilocular. The structure 

 and position of the anthers are very different from that of Ericacece, and bring 

 the genus nearer to Gentianece ; but I am inclined to regard it as the type of a 

 group, alike distinct from these families as well as from Solanece, with which 

 it has also been associated, 



I have confined myself in this paper to the description of the species col- 

 lected by Dr. Royle, who has liberally placed in my hands that portion of his 

 herbarium for this purpose, and some of the more remarkable species will be 

 found represented in his interesting work on the Botany of the Himalayan 

 Mountains. In the arrangement of the species I have adopted some of the 

 divisions of the Linnaean genus Gentiana, first suggested by Renealmns, and 



* My learned friend Sir William .lackson Hooker, in the first number of his interesting and useful 

 work, " Icones Plantarum," has published a figure of what I have long considered to be a third species 

 of this genus, and wliich was first collected by my excellent friend Captain Phillip Parker King, R.N., 

 in tlie Straits of Magellan and in the archipelago of Chiloe, and for which I beg to propose the fol- 

 lowing name and character: 



D. fulgens, foliis cuneato-oblongis dentato-spinosis glabris subtiis glaucis : dentibus divaricatis, seg- 

 mentis calycinis oblongis ciliatis, corolla esdyce 5-pl6 longiore. 

 Desfontainia spinosa. Hook. Ic. Plant, t. 33. haud aliorum. 



The three species, although nearly related, are nevertheless essentially different in their leaves, calyx, 

 and in the proportions of their corolla. 



VOL. XVII. 3 V 



