POLTGONACEOUS GBNERA. 31 



the type specimens from Cameron' Pass at an altitude of 11,700 

 feet, collected by C. F. Baker, 16 July, 1896. 



Of all the new species of Bistorta here proposed the types are 

 in my own herbarium ; and doubtless many more species not yet 

 described exist in other herbaria. The characters on which 

 species may be established — those of rootstock, leaf, spikes and 

 especially their bracts — are herein sufficiently indicated, and 

 other students of the group will carry forward the work here 

 begun upon our North American forms. 



The genus is one of many which while but feebly represented 

 in the floras of Europe and eastern North America, exhibit a 

 multiplicity of species in those of both Asia and western America; 

 and I next subjoin a partial list of such Asian species as accord 

 perfectly with the type as to vegetatiye characters and a simple 

 and terminal spike ; taking for their names under Bistorta the 

 same that have been already assigned them under Polygonum : 

 B. BULBIFBKA (Roylc, Trans. Linn. Soc. xTiii. 94), sphaebos- 

 TACHYA and STEKOPHTLLA (Meisu. Monogr. 53, 52), confusa 

 (Meisn. in Wall. PL As. Rar. iii. 53), pekpusilla (Hook. f. Ic. 

 PI. t. 1490). 



Also Asian, and of this genus as to habit and floral charac- 

 ters, but branched above and bearing several spikes : B. spbciosa 

 (Meisn. Monogr. 66), amplbxicaulis (Don. Prodr. 70) and 

 OXYPHTLLA (Wall. Catal. n. 1715); and lastly 1wo species of 

 Asian mountains that are suffrutescent, bear gr xeful spikes of 

 intensely red flowers, are hardy in England and highly orna- 

 mental under cultivation in Kew Gardens: B. affinis (Don. 

 1. c.) and VACCIKIIFOLIA (Wall. 1. c. n. 1695). 



Prom Bistorta, the distribution of which is rather northern 

 and subalpine, the transition to the almost subtropical genus 

 Tracaulon is every way abrupt. No thoughtful and unbiased 

 mind would be likely to regard the two groups of species as 

 members of one and the same genus ; and in the early history 

 of the typical species they were associated with Fagopyrum or 

 else Hdxine rather than with Polygonum ; so even with Linn- 

 aeus at the first. 



