304 BOTANY OF SOCOTRA. 



these are nearly quite obtuse. Working upon such a basis the synonymy of 

 Rottboll's species is as follows : — 



C. conglomeratus, Rottb. (ref. ut supra). 

 C protcinolepis, Steud. Syu. ii. 15 (cum varietate pro parte). 

 C. conglomeratus, var. pumila, Hochst. in herb. Scbimp. Arab. sect. i. u. 810. 

 C. curvulus, Bcklr. in Linnsea xxxv. 541 — [C. protcinolepis, var. Steud. (pro parte)]. 

 C. arcuatus, Bcklr. loc. cit. 542. 



C.falcatus, Nees et Ehrenb. in hort. Berol.; Bcklr. in Liuntea xxxv. 546 = [C. conglo- 

 merates /3 effusus, Boiss. (pnecipue)]. 

 C. pungens, Bcklr. loc. cit. 537 (?) ; 



and its distribution is pretty extensive, — Cairo (Bove n. 281) ; Nubia 

 (Kotschy n. 21, Petherick) ; Kordofan (Pfund. nn. 245, 246) ; Korjonis (Kotschy 

 n. 20) ; Hanish Id. (Slade n. 8); Jeddah (Zohrab n. 13; Schimper n. 810, par- 

 tim) ; Aden (Hooker, Thomson, Perry) ; Beloochistan (Pierce) ; Affghan. 

 (Griffith n. 6146) ; Centr. Afr. (Schweinf. n. 645) ; El Hami (Schweinf. n. 204) ; 

 Dongola (Ehrenb.). These are the localities from which I have seen specimens 

 in Kew Herbarium. 



It appears that Steudel founded his species C. proteinolepis upon Kotschy's 

 Nubian plant, n. 21. That is unquestionably C. conglomeratus, Rottb. But the 

 variety he mentions there is C. conglomeratus, Hochst. var. pumila, n. 810 

 of Schimper's Arabian Herbarium. Here, as I have already mentioned, are 

 two plants both dwarf forms. One is undoubtedly of C. conglomeratus, Rottb. 

 (and is the type of Bockeler's C. curvulus), the other has caused most of the 

 confusion. It is the dwarf form of the species described by Bockeler as 

 C. proteinolepis, Steud. (var. excl.), under which he quotes Kotschy's Nubian 

 n. 21 and Schimper's Arabian n. 301. Now Steudel's species and Kotschy's 

 plant are, I have pointed out, C. conglomeratus, Rottb., which species Bockeler re- 

 cognises, but Schimper's n. 301 is the same as the second dwarf plant of 

 Schimper's n. 810. Bockeler's description of C. proteinolepis fits very fairly this 

 plant, and on every ground I think we may assume this to be the plant he 

 meant, but it is not Steudel's plant, In taking this specific name then, as we 

 nave done in the last species, we assign it to Bockeler. The only part of Steu- 

 del's plant included is a portion of the variety, viz., the second dwarf plant under 

 Schimper's n. 810. Bockeler's C. arcuatus is a slight variation from the dwarf 

 condition of the species which he made a new species C. curvulus, and C. falcatns 

 is a state represented in a few localities (from the three last mentioned I have 

 seen specimens) with longish narrow slender spikes and very long bracteoles. 

 C pungens is probably also a stout form of the species. 



Recently Boissier (Flor. Orient, v. 369) has considered all those forms to 

 which I have referred as variations of one protean species for which he retains 

 Rottboll's nomenclature, and in additiou adds as varieties C. arenarius, Retz. and 



