202 KEHDER. 



This \'ariety differs from the type chiefly in its 3-5-, rarel^>' 7- 

 foliolate leaves with broader, often ovate, entire leaflets; occa- 

 sionally with simple leaves at the base of the branchlets. All the 

 other distinguishing characters given in the original description of 

 F. macropciala can be found in the typical form. In the type the 

 divisions of the calyx are also attenuate and Aery unequal and the 

 longer divisions equal or exceed the tube in length; the length of the 

 corolla varies between 10 and 16 mm.; the shape of the fruit is very 

 variable even in the same locality, and I ha\'e before me specimens from 

 the Grand Canyon with fruits haA'ing a narrow wing, about 5 mm. 

 wide, 2.8 mm. long and rounded at the apex and others with the 

 wing of the fruit 7 mm. broad and only 2-2.5 mm. long and truncate 

 and emarginate at the apex. On many flowering branchlets all the 

 lea\'es are simple, as in Ward's specimen from the Red Canyon Trail ; 

 a sterile specimen collected by Perci\'al Lowell in the Sycamore 

 Canyon has most of the leaves simple and the others with only one 

 pair of small leaflets at the base. 



Fraxinus cuspidata, var. serrata, n. var. 



Fraxinus cuspidata Sargent, Silva N. Avi. VI. t. 260 (pro parte) 

 (1894); Man. Trees X. Am. fig. 605 (pro parte) (1905) tantum 

 quoad plantam depictam. 

 A typo recedit foliolis manifeste serratis OAatis v. ovato-lanceolatis, 

 plerumque 7, rarius 9 v. 5, paribus inferioribus interdum 3-foliolatis. 



Mexico. Coahuila: mountains east of Saltillo, April 15-20, 

 1880, E. Palmer (No. 796, type); San Lorenzo Canyon, 6 miles 

 southeast of Saltillo, April 16, 1905, E. Palmer (No. 536); Sierra 

 Madre, south of Saltillo, April 12, 1906, C. G. Pringle (No. 13742); 

 Saltillo, cult, at Cotton mill, April 5, 1887, C. S. Sargent. 



This variety seems to be restricted to Mexico and is connected with 

 the typical F. cuspidata by intermediate forms; such are A. Chaves' 

 specimen from Valencia Co., New Mexico, V. Havard's No. 69 from 

 western Texas and Pringle's No. 137 from Mexico, all enumerated 

 above under the typical F. cuspidata. 



2. Fraxinus velutina Torrey in p]mory, Not. Reconnoiss. Leaven- 

 worth to San Diego, 149 (1848).'^— Sudwokh, Hep. Sec. Agric. 1892, 



1 Not F. velutina Lingelsheim in Bol. Jahrh. XL. 216 (Vorarh. Monog. Fraxi- 

 nus, 36) (1907) from Yunnan which belongs to the section Ornus and is rehxted 

 to F. chinensis Il()xl)urgh; it is based on Henry's No. 11S93. I propose for 

 this species the nameF. Lingelsheimii, n. noin. {F. velutina Lingelsheim, non 

 Torrey). 



