34 



81. Hosackia rigida Tioiitli. (iiicl. U. pulerula Roiith. and ff. WngMii Gray). 

 Chiaos Mountains (Foley county). In Hot. Calif, i. 136, Dr. Watsou sugtiests 

 that H. pubinda \\i^,ut\\. and 11. WrujMii Gray are but fortns of H. rujida 

 Benth. Botanists will testify to the impossibility sometimes of distinguishiuK 

 tlic-ae species. In Nealley's coUoction there are some remarkable specimens 

 that combine in one plant all the important characters of these three so-called 

 species. Some of the peduncles are short, and others very lonj; ; the calyx- 

 teeth equal the tube or are shorter ; the leaves are from obovate or oblong to 

 narrowly linear. These specimens should be referred to U. pubirnla if the old 

 specific distinctions are to be kc^pt up. After examining; a lurjre series of 

 s])(5cimons, liowever, it seems best to consider them all but as forms of a 

 wide-spread and very polymorphus species, of which H. Bryanti Brandegeo 

 {I'l. llaja Calif. 144) seems to bo but another form. 



82. Psoralea liiiearifolia Torr. & Gray, var. robusta Coulter, n. tar. Whole 



plant, in all its parts, more robust than the type : leaves linear-oblong, 4 to 

 (5'-"' louff, 5 or C)""" wide, thickly black-dotted above and below: llowcrs 

 mostly in clusters of three, distant along the rhachis. —Clarendon (Donley 

 county). Collected by Nealley in 1888. 



83. Psoralea tenuiflara Tursh. Clienate Mountains (Presidio county). 



84. Dalea alopecuroides Willd. Limjiia canon (Presidio county). 



85. Dalea aiirea Nutt. Satita Anna (Coleman county). 



86. Dalea Domingensia DC, var. paucifolia Coulter, n.var. Whole plant more 



hairy: leatlets but three or four pairs and larger: inflorescence becom- 

 ing more or less compact-clustered in the upper axils, and the calyx-tube 

 nearly glabrous, making very prominent the large amber-colored glamls. — 

 / Eio (irande City (Starr county). This is akso I'almer 1049, collected in 



northern Mexico between San Luis Potosi and Tauipico, and referred liy 

 Hellishly to D. Domivoensis DC. The species has been founil in S. Florida, 

 and Mr. Hemsh^y (lHoJ. Central Amer., i. '2:V.>) credits it to Texas and New 

 Mexico, but from what collectors we are not aware. Mr. Nealley's collection 

 brings the first Texan specimens we have seen, 



87. Dalea formo-sa Torr. Chenate Mount.ains (Presidio county 'I. 



88. Dalea frutescens Cray. Devil's Kiver (Val Verde county), and Chenate 



Mountains (Presidio county). 



89. Dalea mollis Benth. Chisos Mountains (Foley county). 



90. Dalea nana Torr. lioma {Starr county). 



91. Dalea pogonathera Gray, lioma (Starr county), and Chisos Mountains (Foley 



county). The Chisos specimens have unusually broad cuneiform leadets. 

 A Mexican species, apparently extending northward only into southern Texas 

 and New Mexico. 



92. Dalea "Wrightii Gray. Chisos Mountains (Foley county). With unusually 



broad bracts. 



93. Petalostemon emarginatus Torr. ^ Gray. Peiia (Duval county). 



94. PetaloBtemon multiflorus Nutt. Corpus Christi. 



95. Petalostemon violaceus Michx., var. tenuis Coulter, «. var. A slender low 



form rarely as uiucli as a foot high, with round or roundish-oblong small 

 often few-tlowered heads on long slender ]ieduncles, and shorter pointed 

 bracts (not equaling the calyx, and hence not very apparent in the head).— 

 Santa Anna (Coieman county). Ai>parently the form referred to in PI. Fendl. 

 under no. 138. The species is an exceedingly variable <me, but the above 

 variety is so distinct in character that it seems to deserve a name and descrip- 

 tion. 



96. Astragalus leptocarpus Torr. & Gray. Brazos Santiago. 



97. Astragalus Nuttallianus DC, var. trichocarpus lorr. A, Gray. Brazos 



Santiago. Apparently very conunou. 



