452 MR. C. B. CLAKKE ON 



pilosa. Sciniaa " incano-caerulea, altero latere circa hilum radiatim 

 sulcata, altero latere longitudinaliter plicato-sulcata " {Hassk, bene). 

 Hab, in montibiis ad 500-7000 ped. alt., etiamque planitie Beugalise 



orientalis. 



Cyanotis, Kunth^ Enum. 



Stamina 6 perfecta. Flores in axillis foliorum congest! aut in 

 scorpioideis racenmlis intra bractcas geminatas nidulantes. Co- 

 rolla monopetala. Capsnla trilocularis. Semina 6, 



1. Cyanotis axillaris, Kunfh, I. c. p. 105; Bth. in Flora Hong- 

 kong, p. 378. — Tradescantia axillaris, Hooob. Flor. Ind. ii. p. 118, and 

 Coromandel Plants, t. 107. 



Folia linearia. Flores in aXillis foliorum, 1-6, subsessiles, cajrulei, albi 



aut purpurei. Semina rugoso-punctata. 

 Hab. in Bengalia ubique. 



2. Cyanotis cristata, KuntJi, I. c. p. 102; Wight, Icones, t. 2082. 

 Tradescantia imbricata, Rooob. Flor. Ind, ii. p. 120, and t. 1130 in 



Calc. Herb, 



Folia ovato-lanceolata. Racemuli scorpioidei, distantes, parum pilosi. 



Semina longitudinaliter striatula, in facie interiore quatuor magnis 



punctis notata. 

 Hab. in Bengalia ubique. 



This species is well separated from all tbe neighbouring ones 

 by the peculiar markings of the seeds. 



3. Cyanotis barbata, Kunth, I. c. p. 104. 



Folia lineari-lanceolata. Racemuli scorpioidei, distantes, plus minus 



lanato-villosi. Semina rugoso-punctata. 

 Hab. in montibus communis; Himalaya, ad 500-8000 ped. alt.; Kha- 



siya, ad 500-6000 ped. alt. 



There are several described plants very near this species, to 

 Avhich I have given the name found in the Calc. Herb., and which 

 is also the name issued from Kew. But I question whether 

 a large reduction of species ought not to be made here. 



C. barbata only differs from C.fascicuJata (Bth., ' Flora Hong- 

 kong,' p. 378) in tliat it has less woolly hairs on the involucral 

 bracts. Hence, according to Thwaites, Enumeration of Ceylon 

 Plants, p. 323, C. barbata ought to be a synonym of C. lanceolata, 

 Wight, Icones, t. 2085. But, on turning to this figure, I find 

 lliat O. lanccolata, Wight, is a species with axillary inflorescence, 

 and so also prove the specimens of C. lanceolata in this herbarium 



(Bot. Gard. Calcutta). 



I suspect that C. barbata and G. fasciculata should be made 

 varieties of one plant, and that then a long string of Wight's 

 Bpecies should be reduced — probably C. Lawiana and C. rosea. 



