15 



changed it to Phaca. Medic proposed several segregations in 1787. 

 (Jlandula, Glottis, Hamosa, Onix, Stella, Tium and Triquetra Aragallus 

 and Spiesia were proposed by Necker in 1790 but without description of 

 species. Steudel proposed Aragus for the Aragallus of Necker in 1840, 

 and Thium in 1821. Steven in 1832 proposed Ammodytes, chondrocar- 

 pus, Cymbicarpus, Euprepia, Glyclphylla, Picraena, Proselias, Psychri- 

 dium, Rysodium. In 1856 he again continued the dislntregration by 

 proposing Ailurischia, Alopecias, Ankylobus, Craccina, Cystium, Euilus 

 Feidanthus, Hedyphylla, Macrosema, Myobroma, Pedina, Philammos, 

 Solenotus. Torrey and Gray published Nuttall's Kentrophyta in 1840 

 ;r!so Homalobus Nutt. Walpers put out Aurosema in 1842. Boissier pro- 

 posed Europhaca about 1840^ Opiz proposed Kirchnera in 1858, also 

 i^iedyphylla. Royle i roposed Podolanthus in 1835. Kegel & Smalh. Pro- 

 posed Didymopelta, Sewerzowia, and Dipelta in 1877. Diplotheca was 

 proposed by Hochst in 1846. Hamaria was proposed by Fourrnier In 

 1868 and also Hypoglottis and Podoclirea. Rafinesque also earned a 

 place in the segregators by Physondra in 1832. Rydberg seems to have 

 tailed to note that this genus antedates Homalobus by eight years. 

 So we shall see a new batch of synonyms in due time. 



Recently Rydberg harking back to the ancient times has resur- 

 rected the old genera and methods of 8egroa;ati'n s i Ions dead throa^rh 

 the good work of Gray and Bentham & Hooker and the Pflanzenfa- 

 milien. Gray demonstrated conclusively the folly of attempting to 

 segregate the species of Astragalus in different genera, and he has been 

 almost universally followed both in this country and abroad. He saw 

 the folly of keeping up the form genera Kentrophyta and Homalobus 

 as well as Phaca which latter had persisted longest. He stil ad- 

 hered to Oxytropis which can be kept up only on the flimsiest grounds, 

 that of an abruptly beaked keel, and he should have added enlarged and 

 most lobed wings. A few species of Astragalus have an obscure boss 

 at keel tip, and even a rudiment of beak, such as A. acutirostri.^ 

 and nothoxys, and others have produced keel as in campestris, atra- 

 tus, etc. with normal wings, and A. Arizonicus with lobed or en- 

 larged wings, and A. calycosus with conspicuously lobed wings and 

 blunt keel. But it seems better to follow Gray in still keeping up 

 Oxytropis than to merge it in Astragalus, for its species as a rule 

 have a general habit somewhat different from Astragalus if we ex- 

 clude the oxytropidoid species, but with them in the genus it must 

 be merged. 



Britton & Brown in their Flora proposed Orophaca in 1897 for 

 the group Triphylli of Gray, without recognizing the very diverse 

 character of its members. There might be some reason in keeping 

 up Orophaca for A. triphyllus alone but to put the sericoleucus 

 group with it spoils it, as these plants are manifestly modifications 

 of the montanus group. 



The proposed genera of Rydberg are mostly the product of his 

 idea that no genus should contain more than six species whatever 

 Nature may have said or done about it, which is rather hard on the 

 Almighty, but where genera and species are governed by botanical 

 inspiration and not study or morphological knowledge this state of 

 affairs make strange bed fellows. 



Taking up Rydberg's genera alphabetically we find Atolophragma 

 jiroposed for a part of the Alpini. In it are A. aboriginum, Forwoodii 

 (aboriginum) glabriusculus, lineare (aboriginum), elegans, Macounii 

 (Labradoricus var.), Shearii (elegans). These belong together but 

 lie also includes A. Arthuri which belongs in the Hamosi. and .\. 

 T^randegei which belongs in the Stri.gulosi. and A. Ibapensis which 

 belongs in the Atrati. He also omits from the genus the other mem 

 bers of the Alpini. 



His next genus is Cnemidophacos, which contains A., argillosus. 

 confertiflorus, flavus, which naturally belong together (as indicated 

 by Gray under the Ocreati) but he omits A. Moencoppensis and soph- 

 oroides, and puts in it A. terminalis (reventoides) which belongs in 

 the Uliginosi. and A. reventus which belongs in the Reveuti-arrecti. 



