248 BOTANY OF SOCOTRA. 



truncatum, interne acntum, testa tenui comosa ; embryo magnus, cotyledonibus carnosis 

 ovatis plano-convexis accumbentibus radicula longioribus, radicula supeva tereti, albumine 

 nullo. — Suffrutex pulvinatus, parvus, ramis congestis, omnino pilis rigidis adpressis vestitus. 

 Folia alterna, subimbricata, anguste spathulata v. obovata, obtusa. StipuUe 0. Flores in 

 axillis sessiles, spicas unilaterales breves formantes. 



A monotyj)ic endemic genus, and one of the most peculiar plants in the 

 whole flora. 



Its most noteworthy features are its depressed Boragineoid habit and 

 alternate leaves, the calyx adnate to the ovary, the gamophyllous tetramerous 

 regular floral envelopes, four equal epipetalous stamens, the two-celled ovary 

 with a solitary pendulous ovule in each loculus, the loculicidally septifragal 

 capsule containing a single exalbuminous seed embraced by the isolated 

 septum, and the large embryo with accumbent cotyledons and superior 

 radicle, — altogether, making a combination with which I can find no parallel. 



The gamopetaly, andrcecial characters, and the bicarpellary ovary, indicate 

 its position in the series Bicarpellatae of the Gamopetalse, but the somewhat 

 inferior ovary is a character exhibited only in a few Apocynacese, Asclepiadaceae, 

 and Gesneraceae in this series, and with none of these families has our plant 

 any near affinity. 



Amongst the Inferos of the Gamopetalse, the only family in which the 

 technical characters are such as to admit of our considering the question of in- 

 corporating our plant, is Rubiaceas. But the alternate leaves, the absence of 

 stipules and of a floral disk, at once exclude our genus. 



There is no polypetalous order occasionally exhibiting gamopetaly of 

 which we can regard our plant as an aberrant type. 



In spite of the slight epigyny we must, I think, regard our plant as one 

 of the Bicarpellatre ; and whilst I have, for the present, preferred to locate the 

 genus as an anomalous one of the Gamopetalae, I may here indicate the affinities, 

 and the probable ultimate position of it. 



As I have above mentioned, its habit is thoroughly Boragineoid, indeed it 

 has quite the look of one of the desert species of Heliotropium, or some 

 nearly allied genus. Against its incorporation in Boraginese, the capsular 

 fruit is, perhaps, the greatest difficulty to overcome, and then, whilst it has the 

 characteristically superior radicle of the order, the ovule is pendulous and 

 anatropous, not erect with a superior micropylc as is typical. Besides, the 

 accumbent cotyledons are a feature not described in the family. 



Mr Bentham, who has very kindly given an opinion upon the plant, writes 

 <: it is certainly very anomalous. It seems to me to come nearest to Ver- 

 benaceae, amongst which we exceptionally find (though in very few cases) 

 dehiscent fruits or alternate leaves." The characters which seem most to 

 militate against its position in Verbenacese are, in addition to those indicated 

 by Mr Bentham, the accumbent cotyledons with the superior radicle, —this 



