Rydberg : Rocky Mountain flora 461 



tana. The other specimen is a duplicate of the type of A. mollis, 

 received from Hooker. This is almost identical with A. subplu- 

 mosa Greene, or A. Chamissonis longinodosa A. Nels., except that 

 the involucral bracts are broader, oblanceolate, and abruptly 

 short-acuminate. It represents a plant not uncommon in the 

 Rockies, from northern Wyoming northward. The Arnica that 

 is not uncommon in New England was referred to A . Chamissonis 

 in Gray's Manual, 6th edition, and to A. mollis by Robinson & 

 Fernald in the Gray's New Manual. It has nothing to do with the 

 former. It is related to the latter, but is, I think, distinct enough. 

 It has more affinity to A. amplexifolia Rydb. (A. amplexicaulis 

 Nutt.) and A. rivularis Greene than to A. mollis Hook. It does 

 not have the broad bracts of A. mollis. It should be known as 

 Arnica lanceolata Nutt. A duplicate of the type (if not the ac- 

 tual type) of the last named is found in the Torrey herbarium. 

 What Professor Nelson had in mind as Arnica mollis when pre- 

 paring the manuscript of the New Manual, I can not imagine. 

 In his key he separates it from Arnica subplumosa by the "leaf 

 blades decurrent on the petioles." The blades are slightly and 

 but slightly decurrent in both. His description is very vague 

 and evidently drawn from several species. As synonyms he cites 

 "A. Chamissonis in part, but mostly A. latifolia as to our range 

 (A. latifolia A. Gray, Bot. Calif. 1: 415. 1885; A. tomentulosa 

 Rydb. loc. cit. 28: 20. 1901)." It is true that Gray and others re- 

 ferred A. mollis to A. Chamissonis, but I do not know that it 

 has been referred to A. latifolia, unless by Prof. Nelson. Arnica 

 latifolia A. Gray in the Botany of California comprises A . latifolia 

 Bong., A. Menziesii Hook, (this perhaps not specifically distinct 

 from A . latifolia) and A . diver sifolia Greene ( A . latifolia viscidula 

 A. Gray). None of them has anything to do with A. mollis. 

 A. tomentulosa Rydb. is related to A. Chamissonis Less., but differs 

 in the short pubescence and the broad involucral bracts, rounded 

 at the apex. If the form of the bracts should happen to be a 

 variable character and of no specific value, Nelson's own Arnica 

 rhizomata should be reduced to a synonym of A. tomentulosa, as 

 there are practically no important differences except the form of 

 the bracts and the latter name is nearly five months older. 



Greene's idea of Arnica Chamissonis is also wrong. He states 



