346 BOTANICAL GAZETTE [april 



2368, of which the last has to be taken as the type of 5. Thurberi, have been 

 erroneously attributed by Rowlee to G. Thurher, to whom only the following 

 specimen belongs) ; Horse Head Cruping ( ?) River, October 1850, G. Thurher 

 (no. 9S; G.; "10-12 ft."); ?Pecos County, banks of the Pecos, 1889, iVea//?y 

 (no. :3^2>^ n^-; W.); September 1881, V. Havard (m., f.; W.; ad var. typicam 

 accedens); Brewster County, Rio Grande, south of Chisos Mountains, August 

 1883, V. Havard (m., f.; W.); Val Verde County, Del Rio, along streams, 

 October 18, 1916, E. J. Palmer (no. 11069, f.; A.); Potter County, Amarillo, 

 creek banks, July 13, 1917, E. J. Palmer (no. 12539, f-. fr.; A.; ad var. typicam 

 accedens) ; along Rio Grande, near San Vincente, August 26, 191 5, M. S. Young 

 (m., f.; M.); Guadalupe County, in the dry bed of the Cibolo 12 miles east 

 of New Braunfels, August 1851, F. Lindheimer (no. 615 [=1191], f.; G., M.); 

 Comanche County, Comanche Spring, Lindheimer (no. 1190, f.; M.); Mata- 

 gorda County, banks of Peyton Creek near Bay City, May 6, 1916, E. J. 

 Palmer (no. 9689, m.; A.); Cameron County, near Brownsville, November 

 1888, Nealhy (no. 30, f., fr.; W.); (New Mexico?), Rio Grande, July 1848, C. 

 Wright (m.; G.; "small tree"); without locality, 1849, C. Wright (no. 668, m.; 

 G.,W.). 



There have also been described the following forms which I 

 have not yet been able to elucidate: S. longifolia var. sericans Nees 

 V. Esenbeck in Wied-Neuwied, Reise In. N. Am. 2:448. 1841; 

 Engl. ed. by Lloyd, Trav. Int. N.A. 518. 1843, collected on the 

 Missouri, probably in eastern Montana about July 8 (see I.e. 

 1:472 [Engl. ed. p. 211]). I would refer it to S. exigua, but the 

 lower flowers of the male aments are described as "triandri"; 

 otherwise the description agrees with S. exigua. — S. longifolia i. 

 integerrima Kuntze, Rev. Gen. PI. 2:643. 1891, and f. paucidentieu- 

 lata Kuntze, I.e. The first is characterized by the phrase "folia 

 denticulata" and as type is given "U. St., Madisonthal"; while 

 the second has "folia paucidenticulata " and came from " Cheyenne, 

 Nebr." The author adds "Ausserdem kann man eine f. multi- 

 denticulata unterscheiden." I suppose those forms are simply 

 typical 5. longifolia. 



With the hybrids which doubtless occur only too frequently 

 where different species grow together it is impossible to deal, as 

 long as it has not yet been possible to limit the species in a more 

 satisfactory manner. The main purpose of this paper is to point 

 out the correct application of certain names, and to direct atten- 

 tion to such forms as need a close study in the field. 



Arnold Arboretum t 



Jamaica Plain, Mass. 



