INDIAN SPECIES OF UTRICULAEIA. 185 



porrecto crasso cylindrico-conico v. cylindrico obtusissimo labio in- 

 feriore breviore eodemque plus minus obtecto. 



U. racemosa, Wight, I. c. t. 1584. f. 1 ; Benj. in Linncea, xx. 307. — U. 

 cserulea. A, DC. Prodr. viii. 19. 



Loc. In Bengalia, Edgew. Ceylonia, Walker \ Gardner \ (C. P. 2085) 

 Thwaites ! Montibus Pulney, hh. Wight ! 



Radices fibrosse, utriculis paucis nullisve. Folia spathulata v. lineari- 

 spathulata, ante anthesin ssepe evanescentia. Scapus interdum bifidus. 

 Pedicelli brevissimi, bracteam sequantes v. parum excedentes, brac- 

 teolis minutis lanceolatis a basi plus minus volutis. Corolla (pur- 

 purea V. rosea ?) labio inferiore cucuUato striis 4 latis pulchris notato 

 (fide sched. in hb. Hook.) calcar obtegente. Capsula globosa v. sub- 

 globosa, ealycem aequans v. parum excedens, seminibus minutis. 

 U. racemosa (Wall. Cat. 1496) varietas? of. U. complanatam^ Wall. 

 Cat. 1497. 



U. rosea, Edgew., is known to me solely from the description in the 

 * Proceedings of the Linn. Soc' {I. c.) ; I may therefore err in referring 

 to the same form the plant occurring in the South and Ceylon, As 

 noted above, perhaps both may be forms of U. racemosa. Wall. Cat. 

 1496, although I scarcely consider them as such. 



** CorollcB calcar labium inferius excedens. 



I desire especially to draw the attention of Indian botanists to the 

 forms which group themselves under this subsection. Notwithstanding 

 the extensive suites of specimens illustrative of most of these, together 

 with the other aids which have been freely and most liberally afforded 

 me, I feel myself, after devoting no little anxious care to then* study, 

 quite unable satisfactorily to adjust or subordinate them. While, on the 

 one hand, I fear to unite them under one presumed species (the extreme 

 forms of which, however, would by no means present structural differ- 

 ences irreconcilable with the reasonable adoption of such a course), on 

 the other, from the variability of those parts which alone can afford 

 characters not common to the whole series, I cannot, with a fair con- 

 fidence, say between which groups a line may be most suitably drawn. 

 Therefore, although I incline to the opinion that there may be of these 

 two (or perhaps three) species, under existing circumstances I have 

 thought it the safer course to pursue, to indicate merely the groups 

 which I conceive may not improbably be of specific value, without, 

 however, positively attributing such import to them. 



Characteres omnibus communes. 

 Folia lineari-spathulata v. spathulata. Scapus erectus, debilis vel sub- 

 volubilis, floribus paucis v. numerosis. Calyx lobis subsequalibus 

 ovato-rotundatis rotundatisve, superiore saepe obtusiusculo v. subapi- 

 culato corollse labio superiore integro v. emarginato plerumque bre- 

 viore. Corolla labio inferiore cucuUato integro v. subintegro calcaris 



