Blake Revision of Polygala 35 



none or 1-2 pairs, densely or subsparsely puberulous with incurved 

 hairs, 5-17 mm. long, 2.5-8 mm. wide; middle and upper oblong or 

 lance-oblong, obtuse to acutish, 1.1-1.9 cm. long, 3-5 mm. wide, 

 on petioles 1 mm. long. Racemes rather densely flowered, 1.5-4 

 cm. long. Flowers apparently ochroleucous. Sepals lanceolate, 

 pilosulous. Wings oval or obovate-oval, obtuse or abruptly short- 

 pointed, puberulous, 3.5-5 mm. long, 2-2.5 mm. wide. Keel 4-4.5 

 mm. long. Upper petals narrowly obovate, 3.2 mm. long. Cap- 

 sule elliptic, emarginate, sparsely or densely incurved-puberulous, 

 7-7.5 mm. long, 5-7 mm. broad. Seed pilose, 3.5-4.6 mm. long. 

 Aril 1-2 mm. long, distinctly 3-lobed, the lobes scarious, linear or 

 linear-lanceolate, entire or slightly lobulate, descending or the dor- 

 sal more or less spreading, subequal or the dorsal somewhat the 

 longer. Rose! Contr. U. S. Nat. Herb. viii. 315 (1905). P. pu- 

 bescens Gray! ex Chod. Monog. ii. 21. t. 13. f. 33-37 (1893); not 

 Mart. Denkschr. Bot. Ges. Regensb. i. 185 (1815). San Luis 

 Potosi: San Miguelita Mts., Aug. 1877, Schaffner 8 (GK); 1878, 

 Parry & Palmer 42 1/2 (cotype coll. of P. pubescens: GK). 

 Hidalgo: calcareous soil near Tula, 2075 m., 17 July 1896, Pringle 

 6363 (BGK) ; rocks on mountains, Ixmiquilpan, Sept. 1905, Purpus 

 1428 (G). Jalisco: base of Nevada de Colima, 22 May 1893, 

 Pringle 5507 (G). Puebla: Tlacuiloltepec, 1830-2135 m., Aug. 

 1909, Purpus 3920 (G). Mexico: Mt. Zocoalco near Guadalupe, 

 10 July, Bourgeau 477 (K); Valley of Mexico, Schaffner (K); 

 Schaffner 34 (G) ; lava fields, 2290 m., Valley of Mexico, 25 Aug. 

 1896, Pringle 6425 (type coll. of P. compacta: BGK). Mexico 

 without locality: 1856, Schmitz (B). T. 1. Fig. 24. 



It has seemed best to adopt for this plant the name above used 

 in place of the name P. pubescens Gray, most unfortunately dragged 

 forth from its herbarium obscurity by Chodat. P. pubescens Mart., 

 not mentioned by Chodat, was described without locality in con- 

 nection with three other new species of India, and as a MSS. syno- 

 nym of Rottler's is cited for this as for the other three species it is 

 safe to assume that it also is an Indian species. Accordingly its 

 confident identification with P. violacea Aubl. of Guiana by the 

 Index Kewensis is clearly erroneous, as is confirmed by Marlins's 

 description. In the lack of any definite disposition to synonymy 

 of Martius's species, it is certainly inadvisable to use the same 

 name for the Mexican plant here described. This binomial (P. 



