Mr. J. ]\liers on the Tribe Colletiese. 487 



almost agglomerated : the leaves are very small^ ovate, or oblong, 

 fleshy, nerveless, and obsoletely toothed on then- margin, taper- 

 ing into a short petiole ; they are 1^ line long, 1 line broad, in 

 opposite pairs upon a slender nascent shoot about J inch long. 

 From the axils and the articulations of the spines a short sprout 

 is seen, bearing a few leaves and tu'o or four flowers, which are 

 almost fasciculated ; the pedicels are a full line long ; the cylin- 

 drical tube of the calyx is 1^ line long, with four segments of a 

 quarter their length ; the petals are like small white scales, inter- 

 mediate with the segments, and only a quarter of their length ; 

 the anthers are like those of the preceding species ; the ovary 

 and style are glabrous. The flowers are pale when dried, and are 

 probably nearly of a white colour when fresh*. 



7. Trevoa. 



This genus, first proposed by me (in 1825), was afterwards de- 

 scribed by Sir AVilliam Hooker (in 1830), in the 'Botanical Mis- 

 cellany,^ where it is confounded with Talguenea ; he subsequently 

 (in 1833) suppressed the genus, referring the plant on which it 

 was founded to Retamilia, because of its indehiscent fruit. In 

 Trevoa, however, we have a tree of very hard wood, spreading into 

 numerous thick tortuous branches, and equally tortuous branch- 

 lets, which are very spiny, with an abundant foliage, ofi'ering a 

 strong contrast to the bare Ephedra-like branches of Retamilia. 

 In Trevoa the calyx is persistent, does not enlarge, but remains 

 withered, entire, and membranaceous, at the base of its fleshy 

 drupe ; while in Retamilia the whole of the calyx falls away, 

 leaving its short stipitate torus (or small adnate disk) to support 

 a much larger globular fruit. The form of the calyx is also dis- 

 similar in the two genera, and the petals are larger and more cu- 

 cullate. In the species on which 1 founded the genus, the ovary 

 is imiformly bilocular, producing a small 2-celled fruit, often by 

 abortion 1-locular : another species, which Sir William Hooker 

 had inferred to be identical with mine, has a 3-celled ovary, pro- 

 ducing a larger 3-celled nut, which circumstance induced that 

 distinguished"botanist to merge the genus into Retamilia ; but the 

 characters above mentioned are suflicieut to keep them distinct. 

 There is, however, another distinguishing feature in the shape of 

 the stipules, which in Trevoa are deeply bifid, forming a small 

 fuscous linear kind of wing attached to each side of the petiole, 

 of nearly equal length, by which they are attached to the stem ; 

 and the bases of the opposite petioles are connected on each side 

 by a transverse line across the stem. 



By other botanists Trevoa has been confounded with Talguenea ; 

 but the diff"crences here arc yet more strongly marked : the latter 

 * This plant will be shown in the ' Contribntions,' Plate 39 f. 



