76 BOTANY OF SOCOTRA. 



aliter longe saccata. Stylus longus incurvatus ovario fcriplolongior. Leguvicn sparsim 

 strigosum \-\ poll, longum \ poll, latum suturis ambis incrassatis lateraliter dilatatum 

 subtetraquetrum apice subfalcatum rostratum bispermum. 



Nom. Vern. Sidereh (B.C.S.). 



Socotra. On the hill slopes at an elevation of over 1000 feet. B.C.S. n. 

 370. Schweinf. n. 503. 



Distrib. Endemic. 



A species of very striking character. Its fruit alone completely separates 

 it from any described form, and coupled with its peculiar foliage, supplies an 

 easily observed diagnostic mark. The foliage leaves are somewhat variable in 

 form ; especially is this the case in the terminal leaflets of the base of young 

 shoots ; but all present the curious mottled appearance on the upper surface, 

 due to the distribution of the adpressed hairs, from which I have taken the 

 specific name. These hairs are not uniformly disposed over the lamina, but on 

 each side of the midrib densely clothed and sparingly covered patches alter- 

 nate. The hairy patches extend from the midrib outwards and upwards to 

 the margin of the lamina on each side, but the patches do not always 

 correspond on opposite sides of the midrib, and thus a feathered appearance is 

 produced on the surface. Usually there are two or three densely hairy patches 

 on the smaller leaves ; on the larger ones there may be many. 



The local name for this species is apparently the same as for I. intri- 

 cata, Boiss. 



8. I. G-erardiana, Grah. in Wall. Cat. n. 5486 ; Baker in Hook. Flor. 

 Brit. Ind. ii. 100. 



Socotra. On the higher regions of the hills at an altitude over 2000 feet. 

 Abundant on Haghier, south from Tamarida. B.C.S. n. 463. Schweinf. n. 595. 



Distrib. From Afghanistan eastwards through the subtemperate and 

 tropical Himalayan regions. 



This very showy shrub, with flowers much larger than are common in the 

 genus, has not hitherto been found west of Afghanistan. Our Socotran plant 

 is undoubtedly the species, presenting only very slight differences from the type 

 in the size and form of the leaflets. In the type these are lanceolate and rarely 

 over a half-inch long. In our plant they are elliptic or elliptic oblong, with 

 truncate or emarginate apices, and may be as much as one inch long. 



A named variety — heterantha—oi this species, with smaller and more 

 numerous leaflets, is figured in the Botanical Register, 28, t. 57 as Indigo/era 

 Dosua. This is not, however, the true Indigo/era Dosua, Hamilt. (see Baker 

 in Hook. Flor. Brit. Ind. ii. 102). 



9. I. tinctoria, Linn. Sp. 1061 ; DC. Prod. ii. 224 ; Baker in Oliv. Flor. 

 Trop. Afr. ii. 99, and in Hook. Flor. Brit. Ind. ii. 99 ; Wight Ic. t, 365. 



