ROBINSON. — DIAGNOSES OF AMERICAN E'JPATORIEAE. 277 



the memory of the late Marc Micheli, at one time editor of the interesting 

 exsiccatae of which it forms a part. 



Eupatoriastrum Nelsonii, Greenmau, var. cardiophyllum, Rob- 

 inson & Greenman, n. var. Habitu et floribus formae typicae; foliis 

 Buborbicularibus duplicato-deutatis basi valde cordatis. — Los Pinos, 

 Chiapas, Mexico, 2 June, 1904, E. A. Goldman, no. 1052 (type, in hb. 

 U. S. Nat. Mus., tracing and fragments in hb. Gr.). 



Kanimia nitida, Bak. iu Mart. Fl. Bras. vi. pt. 2, 370 (1876). In 

 the Index Kewensis, ii. 3 (1895), Hooker, f. and Jackson reduce this 

 species to A'. " erythralina" crediting the latter name to Bentham and 

 Hooker, f. Gen. ii. 247 (1873). If the combination had been correctly 

 made in the Genera Plantarum it would obviously antedate Baker's 

 binomial and stand according to the rule of priority under the genus, 

 but this is not technically the case. At the place indicated, Bentham 

 and Hooker, f. merely refer to Kanimia, the plant which they call 

 " Mikania erythralina, DC." The combination Kanimia erythralina, is 

 not made, and indeed there is no assurance that the authors of the Genera 

 Plantarum regarded the transferred plant as a distinct species or that 

 iu naming it under Kanimia they would have adopted the combination 

 K. erythralina. All that their statement conveys is the fact that the 

 generic affinities of De Candolle's plant are with Kanimia rather than with 

 Mikania. There is, however, an added reason why the transfer in the 

 Genera Plantarum should not iu this instance be regarded as equivalent 

 to the creation of the new binomial K. erythralina, for De Candolle's 

 specific name was erithalina, given from a fancied resemblance of the 

 nlant to Erithalis. Kanimia nitida, Baker, is not only the first correct 

 combination accompanied by accurate synonymy, but happily it is also iu 

 accordance with the most rigid priority of the specific name, for it is 

 founded on Eupatorium nitidum, DC. Prod. v. 180 (1836), which has 

 priority of position over Mikania erithalina, DC. 



Carphephorus revolutifolius, DC. Prod. v. 133 (1836). This 

 species, described by the eldest De Candolle, was based upon a specimen 

 sent to him by Sternberg from the herbarium of Haenke and supposed 

 to have been collected iu Mexico. Concerning the plant Bentham and 

 Hooker, f. (Gen. ii. 249) say " C. revolutifolius, DC. 1. c. 133, e Mexico, 

 et C. cordifolius, DC. Prod. vii. 267, e Brasilia, nobis ignoti, certe e 

 descr. e genere expellendi sunt." By Hemsley (Biol. Cent.-Am. Bot. ii. 

 108), C. revolutifolius, DC, is still included in the Mexican flora, but 

 with the comment that both this and C. f triangularis, Gray, are doubt- 

 ful species. At the request of the writer, Mr. Casimir De Candolle has 



