NORTH AMERICAN AMYGDALUS WIGHT 131 



separated by prominent diagnostic characters, the species do form a fairly 

 natural group and the name Amygdalus at once conveys to the mind charac- 

 ters very different from those associated with true Prunus. 



Amygdalus texana (Dietr.) W. F. Wight. 

 Amygdalus glandulosa Hook. Ic. PI. 5:t. 288. 1840. 

 Prunus glandulosa Torr. & Gray, Fl. N. Am. 1 : 408. 1840. 

 Prunus texana D. Dietr. Syn. PI. 3: 45. 1843. 

 Prunus hookeri Schneider, Hanb. Launholz. 1 : 597. 1905. 



Leaves oval to oblong-oval, mostly 14 to 25 mm. or sometimes 30 mm. 

 long on young succulent shoots, 6 to 15 mm. broad, slightly narrowed toward 

 the base, mostly obtuse at the apex, green and pubescent above, grayish to 

 mentose below, the margins conspicuously glandular-serrate; petioles 1 to 4 

 mm. long, pubescent; stipules linear and glandular. Flowers appearing in 

 March with or slightly before the leaves, solitary or in twos, 10 to 12 mm. 

 broad ; pedicels 3 to 4 mm. long, grayish pubescent ; calyx pubescent, the tube 

 about 2 mm. long, the oblong-ovate lobes as long and glandular-serrate; 

 petals apparently white, oblong, 4.5 to 6 mm. long, obtuse at the apex and 

 abruptly narrowed to a very short claw. Fruit rather densely velvety-hairy, 

 apparently maturing in June; stone ovoid 13 to 15 nam. long, 11 mm. broad 

 and 9 mm. thick, usually rounded at the base, pointed at the apex and slightly 

 flattened toward the ventral edge but rounded on the dorsal side, the surface 

 smooth. 



A dwarf bushy shrub with very irregular branches and grayish bark, that 

 of the young branchlets light-gray and strongly pubescent. It occurs only 

 in Texas, where it is found in granitic soils from between Laredo and Bejar, 

 the type locality of Hooker's material preserved in the Kew Herbarium, to 

 Llano and Mason Counties and eastward to Hampstead. 



The specific nam.e glandulosa was used earlier for a different species by 

 Thunberg and the name hookeri was published by Schneider for this reason. 

 There appears, however, to be nothing to prevent the use of Dietrich's name 

 texana, which was doubtless overlooked by Schneider. The original descrip- 

 tion by Dietrich was based on a specimen collected by Drummond but for 

 which the locality is not given. 



Specimens examined: Texas ; Stevens Bend, Colorado River, Lampasas 

 Co., C. S. Mason, 1910. Hoover's Hill, near Kingston, C. S. Mason, Mar. 

 26, 1910. Llano Co., Reverchon, May, 1885. Big Branch, Gillespie Co., 

 G. Jermy. Sandy Plains, Hampstead, Elihu Hall, No. 189, June 4, 1872, 

 San Antonio Wells on the Prairie road between Ringgold Banks and Laredo, 

 G. H. Thomas, Feb., 1857. Kimble Co., I. Reverchon, No. 53, May, 1885. 



