1894.] D. Praia — Some additional species of Convolvulacese. 1 09 



Tlie cymes in this plant are almost umbellate and the leaves, which are 4 by 

 3^ in., are much wider than in TAR. typica, where also the corolla is pure white or 

 white tinged with yellow. The calyx and seeds are exactly as in I. cymosa, but the 

 corolla is considerably larger, and in size and colour agrees with that of Ipomoea 

 umhellata. Mey. {Prim. Flor. Esseq. 99), an American plant with very similar leaves 

 equally deeply cordate, but with an acute sinus and glabrous above very sparingly 

 hirsute below. This latter difference is no greater than exists between different 

 forms of I. cymosa proper, and it is probable that I. cymosa, vae. culta, and I. umhel- 

 lata are but forms of one plant which is only a variety, as Bentham [Flor. Austral. 

 iv., 423) suggests, of I. cymosa. 



It has always been supposed that Wallich's Convolvulus umhellatus, cult, in 

 Hort. Calcutta, was derived from American seed ; it now seems as probable that 

 Wallich's plant was of Malayan origin. 



Dr. Stapf who has kindly examined this plant, doubts very much that it is 

 entitled to varietal rank. He also adds " it is extremely like I. umhellata Meyer, 

 from America, and I cannot find characters to separate them." 



45 h. Ipomoea rubens Ghoisy, Convolv. Or. 81 and BG. Prodr. ix., 

 371. Convolvulus rubens Wall. Gat. 1421. C. glandulosus Ham. in 

 Wall. Gat. 22.52. Lettsomia rubens GlarJce, Flor. Brit. Ind. iv., 195. 



IToRTH Bengal: Rangpur, at Pirganj, Hamilton; Purnea, near 

 Cai'agola, Kurz ! Assam : Jenkins ! Gibson ! Goalpara, Hamilton ! Simons ! 

 Gauhati, Jenhins ! Silhet, BeSilva ! Cachar, Keenan. 



This is, as Choisy says, an Ipomoea not a Lettsomia. M. Choisy does not ajjpear 

 to have seen fruit ; Mr. Clarke says, loc. cit., that he had not seen any. The plant, 

 Mr. Clax'ke adds, has been supposed a Bivea ; its facies suggests an Ipomoea in the 

 vicinity of I. cymosa var. culta, from which however, it differs in having fewer flowers 

 in the umbelliform cymes, a tomentose calyx, a corolla which is whitish-purple ins- 

 tead of dark yellow, and strigose on the plaits externally instead of quite glabrous, 

 as well as in having glabrous in place of hirsute seeds. 



There is no example of Hamilton's Convolvulus glandulosus at Calcutta, at Kew 

 or in the type set of Wallich's Herbarium at the Linnean Society ; what however is 

 evidently, from Choisy's description, the same thing, is represented at Calcutta by 

 specimens collected in Assam (exact locality not stated) by Gibson, and at Gauhati 

 by Jenkins. These specimens have rather larger leaves than any of the others 

 densely velvety tomentose on both surfaces, and closely resembling those of 

 Argyreia Boxburghii. There is however not the slightest difference as to calyx or 

 corolla between these specimens and those which form the type of Ipomcea ruhens 

 so that the separation of a variety lanata, proposed by M. Choisy, appears to be 

 hardly necessary. The Goalpara specimens in Wallich's Herbarium {Convolvulus 

 hifidus. Ham. Wall. Cat. n. 1421/B and n. 142I/C) are identical with those of DeSilva 

 from Silhet (Wall Cat. n. 1421/1) on which the species was founded. Kuxz's Pm-nea 

 specimens have leaves less densely hirsute above. 



The species is evidently very closely related to the next one of which there is 

 not a specimen at Calcutta. Being unable to separate it by Mr. Clarke's description 

 and figure, the writer asked that the two might be compared at Kew where the type 

 of Ipomoea Wattii is preserved. Dr. Stapf, who has kindly made the comparison at 

 Kew writes : — " Lettsomia ruhens Clarke, and Ipomo:a Wattii are very like j but note 



