54 ERYTHEA. 
ences. The achene of EHupatorium is not only relatively 
shorter than that of Coleosanthus; it is almost always mani- 
festly tapering from above or near the middle, down to the 
base. In Coleosanthus it is linear and terete, or at least 
more nearly so a good deal than in Hupatorium, besides 
being longer. And, although the Coleosanthus species, now 
quite numerous, are not of uniform habit, none of them agree 
’ with typical Hupatoriwm in that respect. 
The most important dissertations on this genus, after the 
fifth volume of DeCandolle’s Prodromus, were that of 
' Bentham in the Botany of the “Sulphur” (1844), and that of 
Gray, in the first part of Plante Wrightiane. Both these 
eminent men recognized the priority of Elliott’s Brickellia 
over DeCandolle’s Bulbostylis, rejecting the latter and re- 
storing the earlier name, as they were in duty bound by the 
law of priority. Yet both of them, in maturer years, having 
discovered Coleosanthus to be prior to Brickellia, refused 
now to allow the same law to amend their own error; and so 
they stood by that, and retained the name Brickellia in the 
face of the law. It is so much easier to apply the rules to 
other people’s mistakes than to one’s own. The needed 
corrections in nomenclature were made by Dr. Kuntze, as 
late as 1891; but a few of the species that had been some- 
what recently published under Brickellia appear to have 
escaped his eye. These are: 
C. Cedrosensis. Brickellia Cedrosensis, Greene, Bull. 
Torr. Club, x. 86 (1883). 
C. Knappianus. Brickellia Knappiana, Drew; Pitt. i 
260 (1888). 
C. rhomboideus. Brickellia rhomboidea, Greene, Pitt. ii. 
103 (1890). 
C. euspidatus. Brickellia cuspidata, Gray, Proc. Am. 
Acad. xxii. 421 (1887). 
If there be any obstacle to the retention of Coleosanthus 
in generic rank, I conceive that obstacle to be no other than 
