PHANEROGAMS— PROFESSOR EAYLEY BALFOTJE. 241 



The character by which specific diagnoses have been made, — long branching 

 or compact habit, glabrousness or more or less pubescent vestiture, breadth 

 of leaf, solitary or fascicled stalked or sessile flowers, narrow or broad calyx, 

 — are features which are by no means constant, and in our Socotran plants 

 we have almost every degree of variation in these respects. I doubt, indeed, 

 whether it is possible to keep up, even as distinct varieties, all the old specific 

 forms. 



It appears to me that this widely-spread species may in any one locality 

 vary in two directions, and thus all the forms seem to fall into two groups 

 characterised thus : — 



a. remota : ramulis internodiisque elongatis ; foliis rernotis. 



b. imbricata : ramulis internodiisque brevibus ; foliis approximates subitnbricatis. 



Of the first set, M. Jiliformis, Benth. (sp. typ. ex herb. Gay), and M. 

 microphylla, Benth., south European forms, may be taken as typical, along 

 with the Canary island plant, M. Teneriffce, Benth. (M. terebinthinacea, Webb). 

 The branches may be greatly elongated, often rigid, and are commonly more 

 or less purple, and, owing to the length of the internodes, the leaves are far 

 apart. T*hese are frequently purple, usually glabrous and shining, often very 

 small and somewhat narrowed, never subrotundate, but occasionally they 

 become pubescent. The flowers may be either solitary or fascicled and the 

 calyx may be narrow or wide, one-sixth inch long or under one-eighth inch. In 

 this group we include also the Cretan M. sphaciotica, Boiss., the Abyssinian 

 M. ovata (ex herb. Schimp. Abyss, sect. i. n. 1859), some Abyssinian specimens 

 (ex herb. Roth, n. 508), and also some Indian plants (ex herb. Griffith, n. 3979, 

 sub M. biflora). Our Socotran specimens, n. 213, and Schweinfurth's n. 529, 

 also come into this set, which has thus a very wide distribution. 



The second set is typified in the Indian M. biflora, Benth. The leaves in 

 this group are usually larger, much broader, and often nearly rotundate, and 

 owing to the shortness of the internodes are close set somewhat imbricated 

 when dry, and they may be glabrous or pubescent. The flowers are either 

 solitary or fascicled and sometimes are almost or quite sessile. Into this 

 section goes the true M. punctata, Benth., from Abyssinia and Somali Land, 

 and M. Forbesii, Benth., from Teneriffe ; also almost all the Indian and 

 Arabian plants described as M. biflora, and probably also some forms from 

 Abyssinia under M. ovata (in herb. Schimp. Abyss, n. 12). Our Socotran 

 specimens, n. 631, Schweinfurth's n. 600, belong here. The Indian forms 

 we find are, as a rule, glabrous. The Socotran specimens, like those from 

 Teneriffe, are more or less pubescent. One specimen sent by Schweinfurth is 

 much more so than almost any other form I have seen. The Abyssinian plants 

 are commonly pubescent, and have usually elongated leaves. This group has 



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



