12 BOTANY OF SOCOTRA. 



glaucous leaves, which are a favourite food with camels ; and as the young twigs 

 are so usually devoured, it is rare to find flower or fruit. This form I think may 

 be the same as Schweinfurth's n. 251, although he mentions the leaves on 

 young shoots as slender and long-petioled, whilst our plant is heterophyllous 

 after the type of such Mascarene plants as Pyrostria trilocularis, Balf. fil. 

 ScyphocMamys revoluta, Balf. fil., &c. (see Botany of Rodriguez in Phil. Trans. 

 168 (extra vol. 1879). The leaves on young shoots are long and linear, often 2 

 inches in length, and only x^th inch broad, and subsessile, whilst the mature 

 leaves are oblong, about l^th in. long, and shortly petiolate withacuneate base. 

 Is this, then, merely a plain form of the hill plant, or is it a distinct species ? 



M. angolensis, DC. is a wide-spread and variable plant, and as it occurs in 

 Angola is a tree with leaves not narrowly cuneate at the base but rather obtuse 

 and rounded. I have examined De Candolle's type, and also specimens collected 

 by Welwitsch (n. 969)— which last show heterophylly of the same type as our 

 Socotran plain form— and they are certainly different from our specimens. It 

 may be questioned, however, whether the species as taken up and described by 

 Oliver (loc. cit.) is really one or includes more than one form. The Abyssinian 

 specimens in Kew Herbarium, under the name M. retusa, Hochst., quoted by 

 Oliver as a synonym of M. angolensis, are I think doubtfully referable to it, and 

 with them the Socotran hill plant agrees in many ways, especially in the cuneate 

 based leaves, though ours is a more delicate plant and has smaller flowers with 

 a relatively shorter calyx tube. 



Whether there are one or two species on Socotra, and whether the Abyssinian 

 plant is identical with any Socotran form, must remain for future exploration to 

 decide. Meanwhile, as our specimens are not perfect, I have accepted 

 Schweinfurth's description and nomenclature. 



The Socotran tree is a very graceful one, and is thus described by Wellsted 

 (Journ. Roy. Geogr. Soc. v. (1835) 199 : — " The eshaib tree is remarkable as 

 resembling in its light and graceful form the weeping ash of England. Not- 

 withstanding the slender dimensions of its trunk, and its being always slightly 

 inclined in a direction contrary to the prevailing south-westerly breezes, it 

 appears to be capable of withstanding the full force of a tropical storm. From 

 the great length of the petiole, the leaves hang loose, and are easily shaken by 

 the wind, presenting an appearance similar to that produced by the " light 

 quivering aspen." A more beautiful or tasteful mourner over an urn or tomb 

 than this plant could not be selected !" 



4. CADABA. 



Cadaha, Forsk. Fl. iEgypt. Arab. 67 ; Benth. et Hook. Gen. PL i. 108. 



A small genus of shrubs chiefly found in tropical Africa (but extending to 

 the Cape), Arabia, India, and the islands of the Indian Ocean. One species 



