MR. G. BENTHAM ON EUPUOBBIACE^. 221 



in a large proportion American, and distributed by botanists 

 into three very distinct sections, established by Pohl, with 

 some plausibility, as independent genera : — 1st, Curcas has the 

 petals united in a single lobed corolla ; the species are chiefly 

 American, with two or three from Africa or East India; the 

 section also includes Mocinna of Ortega, a Mexican species 

 remarkable for its succulent branches, small leaves, and sessile 

 clusters of few flowers, but with all the characters of the 

 section : 2ndly, Adenoropium, with the petals free or nearly 

 so, to which, besides several American species, belong all the 

 known African Jatropkce : 3rdly, Cnidoscolus, which I should 

 have been much disposed to maintain as a genus, but that 

 Mueller and Baillon both concur in reducing it to a section 

 of Jatropha ; there are no petals, but the calyx is quite excep- 

 tional in the tribe, hypocrateriform and petal-like ; the spe- 

 cies are all American. 8. Acidocroton is a single Cuban spe- 

 cies, copiously armed with infrastipular prickles, otherwise with 



much of the habit and character of Mocinna, but not so easily 

 reducible to Jatropha. 9. Tritaxis, a genus originally founded 

 by Baillon on a Cochin-Chinese plant (known to me only 

 through the detailed descriptions), of which he well pointed out 

 the affinity with Jatropha, although he has since, in his 'Histoire 

 des Plantes,' followed Mueller in reducing it to Trigonostemon. 

 To my eyes it differs essentially from that genus in inflorescence 

 as well as in the stamens, w T hich are nearer those of Ostodes or of 

 some species of Jatropha, and some minor characters. To the 

 original species I would add the Trigonostemon Cumingii, Muell. 

 Arg. (forming Mueller's section Anisotaxis, with a compact sessile 

 terminal cyme), and a new species, from the peninsula of India, 

 with the loose inflorescence of the original one. I would thus 

 characterize the three: — (1) T. Gaudichaudi, Baill. : foliis dentatis, 

 cymis laxis pedunculatis, staminum verticillis 2 pentandris infra 

 terminalem triandrum : Cochin-china (Gaudichaud). — (2) T. Bed- 

 domi, Benth. : foliis integerrimis v. vix sinuato-dentatis, cymis laxi 

 pedunculatis glabris, staminum verticillo unico pentandro infra ter- 

 minalem diandrum : Tinnevelly, East-Indian peninsula {Beddome). 

 (3) T. Cumingii, Muell. Arg. (Trigonostemon): cymis densissi- 

 mis^essilibus pubescentibus, staminum verticillo unico pentandro 

 infra terminalem triandrum : Philippines (Cuming, n. 1693). It 

 is very probable that a specimen gathered by Kurz in the South 

 Andaman Islands may prove to be a fourth species, with the 



LOTH. JOLRN. — BOTANY, VOL. XVII. 8 



