22 BOTANY OF SOCOTRA. 



2. P. spicata, Arn. in Ann. Nat. Hist. iv. (1839), 91 ; Boiss. Flor. Orient, 

 i. 738; Oliv. Flor. Trop. Afr. i. 146 ; Edgew. and Hook. fil. in Hook. Flor. 

 Brit. Ind. i. 246 ; Wight Ic. t. 510. 

 P. staticocformis, Hochst. et Steud. in herb. Schimp. Arab. sect. 1 n. 940. 



Nom Vern. 'Teyeycha. 



Socotra. Not uncommon. Our plants are from the Haghier hills near 

 Tamarida. B.C.S. n. 371. 



Distrib. Abyssinia, Egypt, and through Arabia to the Indian Peninsula. 

 Also Australia. 



This species, which has a wide distribution in the regions around Socotra, is a 

 very marked one in the genus by reason of its Statice-like habit, a character which 

 is shared with it by all the members of the genus in Socotra. Specific limits 

 are at present notably difficult to define in this genus, and the Socotran plants 

 referred to this species depart considerably from the ordinary type as it occurs 

 on the mainland. Normally this is more or less spathulate, fleshy, glaucous, 

 with a few rigid erect branches rising from a basal rosette and subsequently 

 dividing by pairs but not copiously. Now the Socotran plant is more delicate, 

 more copiously branched, with less fleshy leaves, which are usually setose at the 

 point, and whilst the radical leaves are spathulate, they narrow to the base 

 much more gradually, and the leaves on the branchlets are greatly narrowed, 

 becoming at times quite filiform and very long. I should have had more hesita- 

 tion in referring our plant to this species, but for a specimen of Gay's, in Kew 

 Herbarium, labelled P. staticwformis, var. ramosissima, which is certainly P. 

 spicata, but is very greatly branched, and the upper leaves are much narrowed 

 after the fashion of those in our plant, and it is a link uniting our plant with 

 the type. 



But there is also a plant found by both Schweinfurth and our party in 

 Socotra, not abundantly, which I think it is advisable to refer to this species, 

 but as a distinct variety. In general habit it resembles the Socotran forms above 

 referred to, but is very sparingly branched, in this respect coming near the 

 mainland type. But its leaves are very small and thread-like. The basal ones 

 have almost entirely disappeared from our specimens, and as the branches 

 carry but few, the plant has a very bare look. The primary floral axes too are 

 exceedingly attenuated, though straight and ascending. One of the most 

 striking features is the uniform coloration of the bracteoles. These have not 

 a scarious margin with a midrib tinged with colour, as in the type, but with 

 almost no scarious margin, are throughout of a fawn or brown tint, and this 

 contrasts in the dry state very markedly Avith the white scarious sepals of the 

 flowers, which in the fresh condition are deep purple along the midribs. These 

 (lid'crences are evident enough, but as I have, failed to obtain any distinctive 

 characters in the flowers, and moreover, as coincident with the narrowing of the 



