GENERA OF NORTH AMERICAN FABACEAE V. ASTRAGALUS AND RELATED GENERA

Abstract

Linnaeus in the first edition of the Species Plantarum admitted two genera, Phaca and Astragalus, the former containing 3 species, the latter 33. Turning to the fifth edition of his Genera Plantarum, which was published a year later and which better illustrates Linnaeus' concepts of genera than does the Species Plantarum, we find that the most essential character which separates Phaca from Astragalus, is the i-celled legume of the former and the 2-celled one of the latter. The partition dividing the pod of Astragalus is formed by the invagination of the edges of the valves at the lower suture. In certain species, regarded by some as belonging to Phaca, the lower suture is infolded, forming a deep sulcus, but in Astragalus the development has gone a step farther so that the backs of the two valves have fused along the lower edge of the pod, thus forming a double-walled septum. In those species, which the writer regards as representing the genus Astragalus proper, this septum meets the upper suture of the pod and fuses with the same. This is the case in most of the species included in Astragalus by Linnaeus. In a few of them the septum does not extend to the-upper suture and forms only a partial partition, as for instance in A. alpinus. Such species have been referred by some of the later authors to Phaca. In others the partition is formed by the intrusion of the upper suture instead of the lower. These constitute a part of Oxytropis DC. In some species of Cystium, the partition is formed in small part by the upper suture also. In the genus Astragalus, as treated by Bunge, Boissier, Bentham and Hooker, Gray, and later by Taubert in the Pflanzenfamilien, where even Phaca L. and Tragacantha Tourn. are merged in it, the pod has been allowed such an unlimited variation that even Homalobus Nutt. and Kentrophyta Nutt. have been included. These two genera do not even agree with the characters given by Bentham and Hooker or by Taubert as distinguishing the subtribe ASTRAGALANAE. Using their keys or synopsis, anyone would place these genera in the ROBINIANAE. The first segregation from Astragalus was done by Necker in I780, in his Elementa, where Aragallus and Spiesia were published. The former was segregated from Astragalus by the subacaulescent habit, and based on 12 species constituting one group of that genus in the I4th edition of Linnaeus' Systema (edited by Murray). There is not any doubt as to what constituted Necker's concept of the genus, but it is impossible to determine which 584