LEGUMINOS^E. 413 



apex, many-seeded, one-celled, but with both sutures projecting a little 

 into the cell ; the dorsal one not externally appearing introflexed, but 

 the section shows a distinct infolding of the coriaceous endocarp. 



The specimens of the variety, if such it be, are in advanced fruit, 

 and show the root. Their stems, foliage, &c, are similar to those 

 of the species to which it is annexed, only rather more glabrate. 

 But the legumes are larger, from 5 to 7 lines long, and straight or 

 nearly so, less pointed ; the sutures very slightly introflexed; but the 

 projection of the dorsal one is still observable. Additional specimens 

 are requisite to determine whether they belong to a distinct species. 



Specimens of a nearly allied species, if not the very same, exist in 

 the Hookerian herbarium, from Uruguay. This adds another to the 

 numerous species, of various sections, which so invalidate the distinc- 

 tion between Phaca and Astragalus as to compel their union, although 

 the vast number of known species would render the separation 

 extremely desirable, if that were longer possible. Mr. Bentham has 

 already combined the two genera in his as yet unpublished account 

 of the Indian Astragali; and I have followed his example in the 

 second part of Plantae Wrightianas. 



2. Astragalus (Phaca) c-chroleucus. 



Phaca ochroleuca, Hook. & Am. Bot. Misc. 3, p. 186; Gay, Fl. Chil. 2, p. 95. 



Hab. Chili, near Valparaiso. (In fruit only.) 



3. Astragalus (Phaca) canescens. 



Phaca canescens, Hook. & Am. Bot. Misc. 3, p. 185 j Gay, Fl. Chil. 2, p. 105. 



Hab. Chili ; with the preceding. (Without flowers or fruit.) 



This specific name may be retained, since the A. canescens of 

 Solander, published by Mr. Lowe, has been ascertained by Mr. Webb 

 to be identical with A. brachyceras, Ledeb., and no more than a variety 

 of A. hamosus. 



104 



