SmGRB&ATES OS EHUS 116 



A recension of the species of Toxicodendeon is no easy 

 task; so far from easy, I find it one of the most diflBcult I have 

 hitherto undertaken. The best treatment of the genus extant, 

 as to the early and typical species, is that of Dillenius in 1732. 

 Linn^us twenty years later, as his custom was, reduced the 

 genus to Rhus and confused the species. Philip Miller sixteen 

 years after this restored the genus, and also the Dillenian 

 species of it, adding excellent descriptions of two or three new 

 ones. In these two classic revisions of Tournefort's Toxico- 

 DEKDEOK, and not at all in Linnaeus, lie the means of identify- 

 ing all the species early recognized. 



The following represents my present understanding of the 

 names and principal syonymy of the known species. 



T. VULQAEE, Mill. Dict:{17'68); Moench, Meth. 73 (1794.) 



Hedera trifolia Canadensis, Oornut. Canad. 96. Toxicoden- 

 dron vulgare latifolium. Dill. Elth. 389 (1732). 



Khus radicans, Linn, in part, excel, vars. /3 or ;- ; Small, PI. 



727 in part. 



This type species of the genus will have to rest, in the future, 

 as it did with Tournefort, with Dillenius, with Linnseus and 

 and with Philip Miller, on Oornut's Hedera or trifolia Canaden- 

 sis, of which the Cornutian description is fair, and the figure 

 excellent. According to all the authors down to and including 

 Miller, it is a shrub that is often upright and rootless above 

 ground, but sometimes fixing itself to rocks, walls and fences, 

 though never climbing high on trees ; its leaflets ovate, perfect- 

 ly entire, glabrous, or very nearly so ; always with a large fruit 

 and this peculiarly depressed-globose, being distinctly broader, 

 even by its least diameter, than high. This last character is 

 clearly brought out in Cornut's plate, though I am to be the 

 first to mention this mark ; and there is no other species in 

 which this fact holds. 



Miller seems to have declined to adopt for this the Linnsean 

 name radicans. There were two reasons for this course. The 

 Linnsean "species" was an aggregate of three or more; and 

 Dillenius' name vulgare had priority in its favor. In Miller's 

 early day they had not learned that the law of priority was a 

 dead letter anterior to the year 1753. 



