262 University of California Publications in Botany [VOL. 9 



that it was possible to keep A. glabrum Torr. and A. Douglasii Hook, 

 separate; ultimately he reduced to the synonymy of A. glabrum all 

 of the segregates which have been proposed since the publication of 

 that species and also considered A. tripartitum Nutt. (in T. and G., 

 Fl., vol. 1, p. 246. 1838) as impossible of definition as a species. He 

 proposes to recognize two varieties : var. monophyllum Graf von 

 Schwerin (in Gartenfl., vol. 42, p. 650. 1893), with the leaves lobed but 

 not divided, and var. tripartitum (Nutt.) Pax (Engler's Jahrb., vol. 7, 

 p. 218. 1886). This scheme recognizes the distinctness of the Rocky 

 Mountain shrub, which has its foliage normally 3-foliate, a geo- 

 graphical variety which Professor M. E. Jones perceived and expressed 

 in the same combination some seven years before Pax's revision of 

 the family appeared (see Jones, Proc. Calif. Acad. II, vol. 5, p. 627. 

 1895). But this twofold division among the forms of A. glabrum places 

 our Sierran shrub with the variant common in the Northwest; this 

 form of Washington, Idaho, and adjacent Oregon has the leaves 

 shallowly lobed, the angles between the median terminal lobe and the 

 lateral lobes approximating right angles. This appears to be a good 

 geographical race with a fairly well delimited area and to express its 

 relationship to the larger specific unit, Professor C. V. Piper 114 pro- 

 posed the combination A. glabrum douglasii (Hook.), but this shrub 

 most nearly conforms to the plant described by Dr. Torrey as the 

 species A. glabrum. If this species is to be retained, and it must be 

 since the priority of the name is unquestioned, the variant of the 

 Northwest should be so known. Our Sierran form differs from this 

 type species in having the leaf far more deeply lobed, the angles being 

 reentrant, the sinuses sometimes so produced as to give the leaf an 

 appearance not unlike that of the var. tripartitum in extreme cases. 

 This Sierran variant Dr. Greene described as A. torreyi, but both its 

 characters and geographical position indicate that it should be 

 regarded as a var. Torreyi, coordinate with the other two varieties. 



In addition to these three varieties, which express the horizontal 

 distribution of A. glabrum Torr., there is also a well-marked high 

 mountain variety, distinguished by greatly reduced leaves, short spin- 

 escent branches and very white bark, and which of all the segregates 

 proposed, would have, in my opinion, the best claim to be regarded as 

 a species, but which is united to the lower altitude varieties by 

 numerous intergrades. This nearly alpine shrub A. diffusum Greene 

 was described from collections made ' ' near the summit of the West 

 Humboldt Mountains," but is now known to grow in the Sierra 

 Nevada. 



