10 



have had also A. Oreganus, terraiiialis, and A. Cooperi (neglectus) 

 which later he curiously places in Phaca. The allied species A. 

 arrectus, conjunctus, etc. had not then been discovered. This group 

 is evidently allied to the Sarcocarpi. 



His sixth section is Onobrychoides, the Onobrychoidei of DeCan- 

 dolle. It contains A. adsurgens (nitidus) and Hypoglottis (agrestis) 

 two very distinct and little related species but which are closer to- 

 gether than to any other group, and which are probably nearer to 

 the Chaetodontes. 



His seventh section should have been his first, the Reflexi and 

 should not have been separated from the Didymocarpi to which its 

 single species A. reflexus belongs along with A. Brazoensis. 



The eighth section is the Didymocarpi, and ninth the IVIicranthi 

 showing that Gray had probably worked out the relationship of these 

 plants. 



The tenth section (and following sections) takes up another 

 wholly different set of species, of lower development for the most part. 



Section eleven is the Succumbentes, with the one species A. sue 

 cumbens, a class by itself. This may have some relationship with the 

 Galegiformes which he places next but I think is much nearer the 

 Malaci. 



Next comes the Galegiformes two species which seem so radically 

 distinct and yet cannot be separated or placed raiywhere else to 

 better advantage, A. Drummondii and racemosus. To this we now add 

 A. scopulorum. This group is well placed next to the Ocreati, b.it 

 Gray should have placed it in Phaca. The Galegiformes contniiis 

 also A. cyrtoides (A. Gibbsii), Gray guessing from the flowers that 

 it might belong here, but the guess was a poor one as it belongs 

 elsewhere. 



The Ocreati, section eleven, contains three very diverse species. 

 A. flavus is wholly 1-celled, as are the other species we now add to 

 it. So is A. humistratus which belongs elsewhere. The other species 

 A. Oreganus belongs with A. Canadensis. This group is very closely 

 allied to the Bisulcati which Gray places rightly in the Phaca series. 

 This group should also have contained A. gracilis and microlobas 

 which he puts in section 13 and calls the Microlobi. 



The next section the Oroboidei is a hodge-podge of species, a 

 catchall, and shows haste and little study. The first group in this 

 section containing A. aboriginum, Robbinsii, oroboides (elegans) and 

 alpinus (andinus) belongs to his Inflati under Phaca and forms a part 

 of my Alpini, but Gray had to put them here because of the intrusion 

 of the dorsal suture, which however is rot as great as that of the 

 ventral. A. tener he also places here, hvt it belongs among the Lep- 

 tocarpi. In another group in this section he puts A. Lindheimeri 

 and distortus which belong with the Leptocarpi and Hamosi in si)ite 

 of the septum of distortus being slight. He places here A. obcordatus 

 which belongs with A. sparsiflorus, and A. glaber which belongs with 

 the group Hamosi. 



The next class the fifteenth is the Lotiflori containing lotiflorus 

 and villosus, two quite diverse species which have some similarity 

 in flowers only. I think A. villosus is nearer the Sparsifiori to which I 

 refer A. obcordatus and distortus, while A. lotiflorus which is both 

 an Astragalus and Phaca according to the intrusion or absence of 

 intrusion of dorsal suture belongs in a group with A. circumdatus, 

 procumbens (accumbens) and Mohavensis, species not known then. 

 Both these groups are related to the Inflati. Doubtless Gray 

 placed these two species together because they could not be placed 

 elsewhere and had a distant relationship. 



The last class of Astragalus proper, the sixteenth, is the Argo- 

 phylli. This group Gray was also led to confuse by too much emphasis 

 on the intrusion of the dorsal suture, and by too little knowledge of 



