ROBINSON. ALOMIA, AGERATUM, AND OXYLOBUS. 483 



A. viscosum Ort. Dec. 36 (1797) = Stevia salicifolia Cav. 

 A. Wrightii Torr. & Gray ex Gray Proc. Am. Acad. i. 46 (1848) = 

 Trichocoronis Wrightii (Torr. & Gray) Gray, PI. Fendl. 65 (1849). 



3. Revision of the Genus Oxylobus. 



The claims of Oxylobus Moc. to generic rank rest quite as much 

 upon habit and pecuHar habitat as upon readily stated technical 

 distinctions, yet they seem adequate. This small group of three well 

 marked and obviously related species differ from all the Ageratums in 

 being high alpine plants with thickish evergreen leaves. If a very 

 doubtful report of one of them from Venezuela is excepted, they are 

 confined to certain of the highest mountains of southern Mexico being 

 found about at the timber line in borders of coniferous woods. The 

 commonest of the three, 0. arbidifolius, first described by Kunth as an 

 Ageratum, was placed in Phania by DeCandoUe, but is clearly distinct 

 from that genus by the presence of a well developed apical appendage 

 on the anthers. DeCandolle associated with it a second supposed 

 species, his Phania trinervia, known to him only from one of Mocino's 

 sketches. This second species has long remained vague, though 

 Hemsley, Biol. Cent.-Am. Bot. ii. (1881), surmised its probable iden- 

 tity with the earlier species of Kunth. To the writer it seems that this 

 identity may now be stated with definiteness. The tracing of Mocino's 

 sketch, reproduced in the Caiques des Dessins, shows no difference 

 which may not be readily explained by the crudeness of the draftsman- 

 ship. The portion of Mexico in question has in recent years been dili- 

 gently and effectively explored by se^■eral highly trained collectors 

 without the discovery of any second species of just this hal)it. Finally 

 it ma}' be remarked that the eldest DeCandolle appears to have known 

 the Mexican flora chiefly from the collections of Mociho & Sesse, 

 Lagasca, Mendez, Nee, and Haenke and on several occasions rede- 

 scribed unconsciously the species of Kunth, seeming never to have had 

 adequate opportunities to examine the series of plants collected by 

 Humboldt and Bonpland. 



The type of 0. gland iilif ems presents a curious confusion of more 

 technical than practical importance. The species was first distin- 

 guished and named, though not described, by Schultz-Bipontinus, who 

 on herbarium sheets noted a typical bluish-flowered form and a white- 

 flowered condition, which he labelled var. alhifiorum Sch. Bip. It was 

 this white-flowered variety which Hemsley later described as Ageratum 

 gland ulife rum Sch. Bip. and which accordingly becomes ipso facto the 



