580 
umbel, thicker leaves and much larger sessile fruit with longer 
styles. 
With Sanicula gregaria the species has no close affinity. The 
minute campanulate calyx of gregaria separates it sharply from all 
our species, while its 5-divided thin leaves, umbellately branched 
stem and branches, trifoliolate involucral leaves and much smaller 
longer-styled fruit distinguish it radically from Small. The 
branching of Salli is more like that of Sanicula trifoliata than — 
any other one of our species, and selected leaves of the two may be 
closely matched in general shape, but the larger and thinner leaves 
of tvifoliata and its oblong short-styled fruit with the remarkable 
arrangement of the oil-tubes denote its wide difference. 
Sanicula Smalli differs from all other eastern species by styles » 
of medium length, frequently styliferous sterile flowers and thick- 
ened roots. 
In point of this new southern Sanicula I have carefully read 
Rafinesque on his alleged new genus and species “ 77iclinum odora- 
tum,” published in 1817 (Fl. Lud. 79). Taken by itself, Rafinesque’s 
specific description would apply with tolerable exactness to our 
new species, but his generic definition and added remarks, insisting 
on long styles recurved to the base of the ovary, positively ex- 
cludes this new species and, strictly taken, all other known eastern 
species as well. Allowing for possible exaggeration, however, it 
may be well here to take the opportunity of considering further 
our long-styled species Marylandica and gregaria. The specific 
description alone scarcely bars Marylandica unless by the terms 
“folius longe-petiolatus,” which is distinctly not the case with the — 
stem leaves of this species. The same objection holds against 
gregaria, while the further characterization of “ foliolus lateralibus 
bipartitus,” untrue for gregaria, describes exactly the condition in’ 
Marylandica and nearly that ia Smallii. On the other hand, there 
are points in the description which apparently could have applied ie 
only to gregana. In perhaps the most important of these, how- a 
ever, viz. “calyx urceolatus 5-fidus, Rafinesque is himself contra- é a 
dictory, for he concludes by declaring that his genus “differs from 
Sanicula by divided calyx * * *!” Here is a distinction abso- 
lutely without a difference, and it would actually seem to require US _ 
to believe that Rafinesque’s conception of the genus Sanicula as 
