11 



the species. A. pubentissimus was an obscure plant in his day and 

 remained so till I studied it in the field and collected abundant 

 material of it many years later. We now know that it belongs 

 with the Inflati, his Inflati, and not at all with the Argophylli. He 

 also placed A. Arthu-Schottii here, but we now know that this is 

 only a variety of A. lentiginosus and also belongs with the Inflati. 

 The rest of his Argophylli form a well defined group which must 

 be recognized but must be extended to contain some of his Phaca 

 group, and whose relationship to his Argophylli he clearly saw as 

 is shown by their position next to the Argophylli. The Argophylli 

 then should contain not only the original species A. Missouriensis, 

 ohortianus, Parryi and glareosus (erroneously supposed by Gray 

 to be A. argophyllus) but also his Phaca species A. inflexus, the 

 true glareosus, Purshii and Utahensis. Very close to this group, 

 as we have stated before, is the Mollissimi, and various species discover- 

 ed since Gray's time. 



Taking up series 2, PHACA, we find it subdivided into 11 groups. 



The first group the Eriocarpi does not belong in Phaca at all 

 as I have stated, but belongs in Argophylli. 



The second group (Section 18) the Oocarpi contains A. Cooperi 

 (neglectus) and oocarpus, two quite diverse species. The first be- 

 longs with A. Canadensis as I have stated and is an Astragalus, 

 and the second belongs in the Inflati. A. Preussii he places in the 

 Scytocarpi and which group is next of kin to the Scytocarpi on the 

 one hand and the Inflati on the other. 



The third group (section 19) is the Inflati, a well defined group 

 but contains A. frigidus (alpinus) which as I have shown belongs in 

 his Astragalus along with A. aboriginum- in a group which I have 

 named the Alpini. 



The fourth group (Section 20) is the Lonchocarpi with one species 

 A. lonchocarpus which is an Astragalus and not a Phaca, having the 

 ventral suture not at ?11 produced and belongs next to the Galegiformes. 



His fifth group Section 21) is the Microcystei, all of whose 

 species belong in the Inflati. 



The sixth group is the Bisulcati to which he refers A. Beck- 

 withii and bisulcatus, two species not at all related. He might much 

 better have referred A. Beckwithii to the Sclerocarpi which ?t 

 resembles in many ways, but extensive field work has shown that it 

 is closest related to A. oophorus Avhich belongs to the Inflati of Gray 

 but connecting with the Preussii. A. bisulcatus is so closely related 

 to the Ocreati that it hardly deserves a separate grouping. . 



His seventh (Section 23; group is the Pectinati containing A. 

 pectinatus only, but which I think is better placed among the Podo- 

 sclerocarpi. 



His eigth group is the Scytocarpi, another botanical catchall 

 of no merit. Its species are nine. A. Preussii belongs with the 

 Preussii, A. Chamaeleuce (pygmaeus) with the Argophylli as also does 

 A. tephrodes. A. aridus belongs in his Inflati. A. Sonorae belongs 

 with the Homalobi. A. gracilentus, Hallii, Fendleri, and flexuosus 

 belong together in p group I have called the Flexuosi, and which 

 ranks lowest in th° g'-nus along with the Homalobi. 



His ninth group (Section 25) is the Podo-scierocarpi containirT 

 A. sclerocarpus and speirocarpus, which I think belong in separat-^ 

 groups, A. sclerocarpus represents a well developed group of desert 

 plants containing his Pectinati, Watson's Pterocarpi, and several 

 other species. A. speirocarpus belongs with A. collinus and Gibbsii. 

 Tweedyi, _ Alvordensis, and porrectus in a separate group which T 

 have called the Collini and which is nearest related to the Flexuosi 

 on the one hand and the Podo-sclerocarpi on the other. 



His tenth group (Section 2G) is the Homalobi which he pretty 

 clearly defines, though I would place A. collinus in a separaie group. 



