Astragalus vexilliflexus var. vexilliflexus

Barren clay bluffs, rocky knolls, and outcrops of shale or sandstone, ascending in open places to rocky slopes, crests, and talus slides in the mountains, 3100-8000 feet, locally plentiful in the northern Rocky Mountains from Banff National Park, Alberta, south, mostly east of the Continental Divide, to the big Horn Mountains, the lower Wind River Valley and Yellowstone Park in Wyoming, east to southern Saskatchewan, the Big Snowy Mountains in Montana and, apparently greatly isolated, to the badlands of South Dakota (Pennington County), west in southwestern Montana to the upper forks of the Missouri and just across the Divide to the head of Deer Lodge Valley. — Map No. 35. — Late May to August.

Astragalus vexilliflexus (with bent banner) Sheld. in Minn. Bot. Stud. 1: 21. 1894, a legitimate substitute for A. pauciflorus (few-flowered) Hook., Fl. Bor.-Amer. 1: 149. 1831 (non Pall., 1800).—"Among rocks in the more elevated regions of the Rocky Mountains, Drummond."—Holotypus, K! isotypi, labeled "Drummond, misit Hooker," GH, NY (fragm. in herb. Torr.)!—Tragacantha pauciflora (Hook.) O. Kze., Rev. Gen. 947. 1891. Homalobus vexilliflexus (Sheld.) Rydb. in Mem. N. Y. Bot. Gard. 1 (Fl. Mont.): 249. 1900.Astragalus amphidoxus (doubtful) Blank, in Mont. Agric. Coll. Sci. Stud. 1: 72. 1905.— "Sky High, Unionville, 6000 ft., July 10, 1898, E. N. Brandegee."—Holotypus (MONTU ?), not examined; isotypus, NY! Unionville is a small community a few miles s.-w. of Helena, Montana.—Homalobus amphidoxus (Blank.) Rydb. in N. Amer. Fl. 24: 264. 1929.

As described in these pages, the typical variety of the bent-flowered milk-vetch, var. vexilliflexus, is a polymorphic aggregate of two or possibly three minor entities which may deserve taxonomic recognition as they become better known. However, it does not seem possible at present to formulate any clear differences between them. In the northern Rocky Mountains, from the limit of the species in Alberta south at montane elevations (mostly over 6000 feet) to the Little Belt and Bridger Ranges in Montana, the prevailing form of var. vexilliflexus is a prostrate, more or less densely matted plant with relatively large flowers (banner ± 7-9 mm., keel 4.5-5.5 mm. long) of a lively pink or pinkish-purple hue. The foliage tends to be neat and small, the internodes short, and the whole plant is usually quite thinly pubescent, the foliage only rarely becoming cinereous or silvery. It is the original Astragalus pauciflorus of Hooker. In the foothills and intermontane valleys of western Montana, from the Little Belts south to the Yellowstone and Ruby Rivers and Yellowstone Park in Wyoming, var. vexilliflexus is represented by a more loosely and diffusely branching plant with somewhat smaller flowers (banner mostly 5-7.5 mm., keel 3.5-4.5 mm. long) varying from pink-purple to whitish tipped with lavender. The internodes tend to be longer than in the montane phase, while the foliage is a little more ample and commonly ashen. The racial situation in Montana is well illustrated by comparison of the series of specimens collected by Hitchcock and Muhlick in the Little Belts (Nos. 12,223, 12,359) with their material from the foothills of the Belt and Crazy Mountains (Nos. 12,076, 13,306) not far distant. However, some specimens of var. vexilliflexus from outlying stations (e.g., from Cypress Hills, Saskatchewan) are intermediate in flower-size, and even in Montana occasional examples of variation in petal-length among members of a single colony serve to blur the ideal differences. Both in montane and lowland plants the pod varies from symmetrically to quite obliquely elliptic in profile, and A. amphidoxus, maintained by Rydberg (1929, pp. 257, in clave, 264) chiefly on account of its asymmetric fruit, has no claim to recognition on this score. The typus of A. amphidoxus has small flowers (banner 5.6-6.6 mm., keel ± 4 mm. long) described as violet but now faded, and seems to represent a particularly slender phase of the lowland type, despite its origin at a reported elevation of 6000 feet. It may deserve more attention when topotypes can be secured.

Finally, quite local at low elevations in the Big Horn Basin and Wind River Valley, A. vexilliflexus reappears in a third phase distinguished by its erect, bushy-branching stems and extremely small, white or whitish flowers (banner ± 4.3-6 mm., keel 3—3.4 mm. long) at least sometimes giving rise to tiny pods as little as 3-5 mm. in length (but these seldom observed). The extreme phase (Parry, N.-W. Wyoming Exped. 82, NY; Ripley & Barneby 7971, RSA) resembles some forms of A. tenellus, although it differs in its truly sessile pod. This form probably will merit recognition as a named variety, although we already have suggestions of inter gradation into the lowland Montana form described above. In a preliminary note on variation in A. vexilliflexus (Barneby, 1956, p. 484) it was treated as a minor variant; but this course really evaded the problem of racial differentiation in the species which requires extensive field study for solution.