42 



no imolucre, and involuccls of few short subulate bractlets; rays 1 

 to 8 inches long; pedicels 5 to 8 lines long: fruit oblong, glabrous, 

 8 lines long; dorsal and intermediate ribs very prominent, lateral 

 wings thin, about half as broad as body: oil-tubes 25 to 30 and 

 continuous, 8 to 10 of them on the commissural side: seed-face 

 plane. (Fig- l^'^-) — ArchaugcUca atropurfnrca Hoffm. Um- 

 bel. IBl. 



Low river banks, from Labrador to Dolawaro, and westward to Illinois 

 and Minnesota. Fl. June. 



A. LirciDA L. is referred to Canada by Cornuti, upon whose 

 authority alone it stands as a North American species. It has long 

 been cultivated in Europe, but its existence as a member of our 

 flora is so very improbable that we do not include it. See Torr. 

 & Gray, Fl. i. 621. 



A. VERTiciLLATA Hook. is a Very uncertain species, and 

 judging from the description given (which is the only information 

 accessible) Mr, Watson thinks it probably belongs to some other 

 genus. See Watson Proc. Am. Acad. xvii. 374. 



8. SELINUM Linn. Gen. n. 337. — Tall stout branching per- 

 ennials, with pinnately decompound leaves, few-leaved involucre, 

 involucels of numerous narrow or broad bractlets, and white 

 flowers. 



This genus is considered liard to distinguish from Ligusticuin, ^^^ 

 with our own species there is no such trouble. LiV/»s'<"cmwi is character- 

 ized by its fruit having equal ribs, numerous small oil-tubes, and prominent 

 conical stylopodium, as well as by its leaves being always ternate in their 

 primary divisions. Seliniim, on the other hand, has fruit more winged, the 

 lateral wings being the broadest, prominent solitary oil-tubes, usually de- 

 pressed stylopodium. and mostly pinnately compound leaves. The genus 

 seems more nearly allied to Angelica, through';such a species as SeUnuin 

 Hookeri. The important distinctions are that the dorsal and lateral 

 ribs of Selitmm are decidedly and equally winged, sometimes nearly as 

 much so as the laterals; while the dorsal ribs of Angelica are not at all 

 winged, except such irregular winging as occurs in A- Curtisii- The 

 leaves of Selinum are, as a rule, pinnate; while those of Angelica (with the 

 single exception of A- pinnata, which can hardly be considered a typical 

 Angelica) are ternate, at least in their primary divisions. In Angelica the 

 bractlets, if any, are small and very narrow; while in Selinum they are 

 frequently prominent and sometimes broad. Selimim usually has, also, a 

 less flattened carpel, and always strictly solitary oil- tubes. 



