l8 9 I -J Notes on North America?i Willows. 103 



growth of the individual tree from which Wright's specimens 

 were taken. Nevertheless, as Andersson insists upon the 



" short, thick, densely flowered amcnts," as essentially dis- 

 tinguishing this variety, it may be as well to avoid forcing a 

 decision upon the scanty materia! at present available. S. 

 nigra var. venulosa extends to the Sierra Nevada Mountains, 

 but whether Rolander's specimens (leaves only), from Cache 

 Creek near Clear Lake, show a reversion on the Pacific slope 

 to something like the typical form of the species, or whether 

 these were taken, late in the season, from the extreme tips of 

 branches of var. venulosa, or whether it is not even more 

 probable still that these long, narrow, sharply serrate leaves 

 indicate an outlying northern station for 5. Humboldtiana 

 var. oxypJiylla is a question that can receive no satisfactory 

 answer until much more is known than we now know of the 

 forms which 5. nigra assumes in southern California and 

 northern Mexico. 



2. S. LAEVIGATA Bebb. — This fine willow reaches its ful- 

 lest development in central and southern California. It is not 

 know as an Oregon species, though collected by Mr. Joseph 

 Howell, just south of the boundary line in Siskvyou county. 

 Southward it takes on a serotinous mode of inflorescence, like 

 other northern species which invade the tropics. 



3. S. LASIANDRA Benth. — Local observers may be inclined 

 to regard the more pronounced varieties as distinct, but if so 

 1 am unable to limit them. In a broad and comprehensive 

 view the propriety of uniting the Rocky mountain and Pa- 

 cific coast forms under one species and keeping this distinct from 

 5. lucida of the Atlantic coast will, I believe, be conceded. 

 S. lucida VM\ macropltylla Anders, referred "from the descrip- 

 tion" to X. lasiandra var. lancifolia I have since seen, not 

 only in Dr. LvalTs (type) specimens, but in others from the 

 Columbia River, in which the peculiarity described is exhibit- 

 ed in a still more marked degree. It is simply a broad- 

 leaved, showy-flowered state of var. lancifolia with nothing 

 whatever to indicate any particular affinity with S. lucida. 



4,5, «;. THE LoNGIFOLI^- — This group is distinctively 

 American, clearly defined on every side, shading off into no 

 other by variation, hybridizing with none. It is not con- 

 nected with Old World forms by any synthetic type of the 



present or of any preceding period, but apparently was de- 

 rived from the Mexican plateau at the close of the Tertiary. 



