PHANEROGAMS— PROFESSOR BAYLEY BALFOUR. 209 



Order LIV. ACANTHACE^. 



A large order of the tropical and warmer regions of the globe, rarely found 

 in temperate regions. The order has a remarkable development in Socotra. 

 Twenty-seven species in all are known from the island, and they are referable 

 to fifteen genera. Of the genera, three are endemic, one of them being- 

 represented by three species, and another by two. Of the remaining twelve, 

 one is only known from Arabia, seven are entirely African and Asiatic in dis- 

 tribution two of them reaching to Madagascar, one is a widely-spread old 

 world tropical genus, and the rest are dispersed more or less widely over the 

 tropics and warmer regions of the whole globe. 



1. RUELLIA. 



Ruellia, Linn. Gen. u. 784; Benth. et Hook. Gen. PI. ii. 1077. 



A large genus of often showy plants, having its headquarters in tropical 

 America, but spread throughout warmer regions of both the old and new 

 worlds. Of our Socotran species, two are endemic, and the third is a widely- 

 spread south-west Asiatic and tropical African species which occurs also in 

 Madagascar. 



1. R. patula, Jacq. Misc. Bot. ii. 358, and Ic. PI. Rar. i. t. 119 ; T. Anders. 

 in Journ. Linn. Soc. vii. (1864), 24, and ix. (1867), 460. 



B. matutina, Hochst. et Steud. iu herb. Schimp. Arab. n. 874. 



Dipter acanthus patuhis, Nees ab. Esenb. in DC. Prod. xi. 126 ; Wight Ic. t. 1505. 



D. erectus, Nees ab. Esenb. in Wall. PI. As. Ear. iii. 82. 



Socotra. On the plains near Galonsir and elsewhere. B.C.S. n. 184. 

 Schweinf. nn. 452, 781 in lit. Nimmo. 



Disteib. Indian Peninsula and south-west Asia, tropical Africa and 

 Madagascar. 



All the Socotran specimens of this common plant show small-leaved pro- 

 strate forms, with the flowers solitary in the leaf-axils, and the fruits somewhat 

 smaller than in the type. 



There are three well-marked forms on the island. 



The first (our n. 184, Schweinf. n. 452) has the clothing of the type, and is 

 like some of the Arabian plants, and those of Nimmo in Kew Herbarium. 



The second set of specimens are from an altogether more densely pubescent 

 plant ; very thickly pubescent it is, with rounded, obtuse leaves, and the 

 corollas one-third largerthan in the foregoing form; the style, too, is at least a third 

 longer, and has more expanded lobes. The plant from Nile Land collected by 

 Speke and Grant, referred to this species by Oliver (in Trans. Linn. Soc. xxix. 

 (1875), 127), is probably our form. It may be well described as a variety, as 

 follows : — 



TRANS. ROY. SOC. EDIN. VOL. XXXI. 2 D 



