INTRODUCTORY CHAPTER. xlv 



and compact inflorescence. Both species of Ectadiopsis differ from the type of 

 the genus in the form and relation of the corona-scales, and E. volubilis is 

 unique on account of its twining habit. Vincetoxicum linifolium, another 

 endemic species, has fiagelliform twining branches and narrow linear leaves, 

 features only observable in some West Indian species of the genus. The 

 distribution of Sa?'costemma Daltoni deserves notice. It is known with certainty 

 elsewhere only in the Cape de Verde Islands, and the Socotran plant differs 

 in trifling details alone from the form in these islands ; but it is likely that it 

 ranges over the highlands of tropical Africa, some fragmentary specimens from 

 these regions having the facies of the species, though identification is not 

 possible. We have only been able to identify for certain one out of four 

 Stapeliew from the island, which proves to be endemic. A second is probably 

 the monotypic Abyssinian Echidnopsis. There are in our collection, besides 

 the named forms and those referred to genera but not specifically known, 

 portions of several asclepiadaceous species, and future explorers of the island 

 will be able to increase greatly the number of known plants of the family. 



The Socotran Boraginese are chiefly remarkable because they comprise an 

 endemic genus, Cystistemon, which has beautiful azure-blue flowers, and is 

 well worthy of cultivation. Its affinity in the family is somewhat doubtful. 

 Trichodesma Scotti, another of the endemic species — which are nine in all, or 

 one-half the order, — has this to mark it, — its flowers are the largest of any in 

 the genus. Heliotr opium, as might be expected, includes a large portion, one 

 half, of the species of the order, most of them being scabrid plants of the plains. 

 Cor diet obovata may have mention, because its succulent fruit is one of the few 

 edible fruits on the island, — very poor as a fruit it is. 



In the Labiatse, of which nearly half the species are endemic, there is 

 nothing requiring special mention except, perhaps, the occurrence of two 

 species of Lasiocarys, a genus represented elsewhere in two south African 

 and one Abyssinian species, and the remarkable mimicry between the foliage 

 of Lasiocarys spiculifolia and the acanthaceous Blepharis spiculifolia. 



Scrophularinese in Socotra exhibits many features of interest. Its thirteen 

 species are distributed in twelve genera, Striga being the only genus with more 

 than one species; one-third of the representatives are endemic, and one constitutes 

 an endemic genus. Seven of the species are indigenous annuals, and one of 

 them is endemic. The endemic genus Xylocalyx is very peculiar on account of 

 the increase in size, hardening, and lignification of the calyx after flowering ; 

 its alliances are chiefly tropical African. Graderia fruticosa is an endemic 

 member of a genus hitherto described as monotypic south African, but I have 

 seen specimens of an undescribed Angola species. The endemic Campylanthus 

 spinosus is another interesting geographical species. Only four other species 

 of the genus are known, two being from Cape de Verde Islands, a third known 



