Vol. 8, No. 4 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN October 5, 1953 



A REVISION OF THE GENUS BRACHYOTUM 

 (TIBOUCHINE AE-MELASTOMACE AE f 



John J. Wurdack 



INTRODUCTION 



Brachyotum is the second-largest genus of the tribe Tibouchineae (Melasto- 

 maceae), exceeded in number of species only (and dwarfed) by Tibouchina. All 

 species of Brachyotum are endemic to the Tierra Fria and upper Tierra Templada 

 (Pennell 1951) of the Andes from north central Colombia (Dept. Antioquia) to ex- 

 treme northwestern Argentina (Dept. Jujuy), generally at altitudes of 2000 to 4000 

 meters. The ecologic niches occupied by the various species are: the "paramo" 

 and "paramillo" (subparamo) regions of Colombia and Ecuador; the "monte" 

 ("ceja de la montana," "monte de arroyada," and the lower margins of the "pa- 

 jonales") of Peru (Weberbauer 1945); and the "Ceja" and "Bergwiesen" of Bo- 

 livia and northwestern Argentina (Herzog 1923)* Specific ranges are in general 

 quite limited. Three species are found only in Colombia, two in Colombia and 

 northern Ecuador, sixteen only in Ecuador, twenty-one only in Peru, two in south- 

 eastern Peru and northwestern Bolivia, and one from northern Bolivia to extreme 

 northwestern Argentina. No species has crossed the Olmos break (Pennell 1951) 

 in northern Peru; indeed, the species found immediately on either side of this 

 break, although generally related, are more abundantly distinct than members of 

 some species-pairs with very great geographical disjunctions (cf. B. cernuum and 

 B. grisebachii). 



NOMENCLATURAL HISTORY 



Brachyotum was established as a genus by Triana (1867) to include Arthro- 

 stemma sect. Brachyotum DC. and various species previously ascribed to Chaeto- 

 gastra DC. and Rhexia L. Triana (1871) recognized 28 species in the first treat- 

 ment of the genus. The earlier disposition of the species described before Triana's 

 revision of the family has an involved history correlated with the confusion in 

 generic delimitations up to that time. There are no problems, however, in the no- 

 menclature of any species published by taxonomists before A. P. de Candolle. 



Chaetogastra DC. (Candolle 1828a) included species now assigned to fAeri- 

 ania (Swaitz 1800), Tibouchina including Purpurella (Aublet 1775), Desmocelis 

 (Naudin 1849), Pterogastra (Naudin 1849), and Pterolepis (Miquel 1840; nom. con- 

 serv.), as well Brachyotum strigosum, B. cernuum } B. canescens, and B. con- 

 fertum. To obviate any further use of Chaetogastra DC. (which might cause no- 

 menclatural changes in any of the several genera published subsequently to 1828 

 and whose names have not been conserved), Chaetogastra longifolia (Vahl) DC. 

 is here designated as the type of this genus, since this species is now univer- 

 sally placed in Tibouchina. C. longifolia was placed^by de Candolle in the sec? 

 tion of the genus which he characterized as "les vrais Chaetogastra" (1828b). 

 With this typification, Chaetogastra DC. p.p. becomes a synonym of Tibouchina, 



Submitted in partial fulfillment of the requirements for the degree of Doctor of Phil- 

 osophy, in the Faculty of Pure Science, Columbia University. This work was supported in 

 part by a grant from the Davella Mills Foundation, to which the author expresses his grate- 

 ful appreciation. 



343 



