Low, loosely or densely tufted, with a taproot at first very slender becoming stouter, woody, and ultimately producing a shortly forking caudex, the herbage pubescent, usually densely so, with fine hairs affixed shortly above the base, very variable in length and orientation, the vesture strigulose, pilosulous, or hirsute, composed entirely or in part of appressed, ascending, or spreading, and entirely or in part of straight or sinuous hairs up to 0.5-2.2 (2.5) mm. long, the leaflets cinereous or canescent, the upper surface commonly of a lighter green than the lower and often medially glabrescent or glabrous; stems 1-several, almost 0 up to 7 (12) cm. long, in seedling plants often solitary and erect, in mature ones more numerous, the outer of the tuft decumbent or prostrate, the internodes all short, concealed by stipules or up to 1.5 (3.5) cm. long, mostly shorter; stipules submembranous, pale green, pallid, or purplish-tinged, becoming papery-scarious in age, broadly deltoid, ovate- or triangular-acuminate, or lanceolate, (2.5) 3-8 mm. long, ± semiamplexicaul-decurrent; leaves 2.5-11 (14) cm. long, with slender petiole and 7—17 or in many seedling (and some mature) plants only (1) 3—9 narrowly to broadly elliptic, oblanceolate, oval, or obovate-cuneate, flat or loosely folded leaflets 4-20 (26) mm. long, the terminal one commonly (but not quite always) longer than the last pair; inflorescence dimorphic: a) chasmogamous and b) cleistogamous, these found commonly on different plants or if on the same plant at different seasons of the year, most often in discrete colonies of like plants, but sometimes separately on different individuals in a colony, in seedling plants cleistogamous only; a) chasmogamous racemes borne on slender peduncles (3) 4—12 cm. long, these ascending at anthesis, procumbent in fruit, either longer or shorter than the leaf; flowers (3) 5—17, rather densely racemose, at first ascending and forming an ovoid or subcapitate head, subhorizontal in age, the axis somewhat elongating, 7-25 mm. long in fruits; bracts thinly herbaceous or submembranous becoming papery, lanceolate or ovate-acuminate, 1.8-5 mm. long; pedicels at anthesis straight, ascending, 0.5-1.3 mm. long, in fruit thickened, sometimes arched outward, 1.2—2.4 mm. long, persistent; bracteoles usually 0, occasionally present at or below the base of the calyx; calyx 5-9.7 mm. long, often purple-tinged, the scarcely oblique disc 0.5-1 (1.2) mm. deep, the campanulate tube rounded at base, 3.2-4.5 mm. long, (2.2) 2.4-3 mm. in diameter, the lanceolate, often slightly incurved teeth 2.2—5.2 mm. long, the whole becoming papery, ruptured, marcescent; petals greenish-white, ochroleucous, or tinged or veined with lavender, sometimes brightly tipped or margined with purple or bluish-purple; banner recurved through 45° or sometimes further, ovate- or oblong-cuneate, shallowly or deeply notched, 8.5-14 mm. long, 4.6-6.4 mm. wide; wings 7.8-11.8 mm. long, the claws 3.1-4.4 mm., the oblanceolate, narrowly obovate, or rarely linear-oblong, obtuse, nearly straight or gently incurved blades 5.3-8.3 mm. long, 1.6-3 mm. wide; keel 6.6-9.8 mm. long, the claws 3.4—4.6 mm., the half-obovate blades (3.3) 3.6-5.9 mm long, 2.1-3 mm. wide, abruptly incurved through 90-110° to the bluntly or exactly deltoid apex; anthers (0.4) 0.45-0.65 mm. long; b) cleistogamous racemes borne on short or subobsolete peduncles mostly shorter than the subtending stipules, occasionally up to 2 cm. long, the fruits in consequence subradical and concealed beneath the tufted foliage; flowers 1-3 (5), very shortly racemose; calyx at anthesis commonly shorter, but accrescent and as large in fruit, the tube at first narrower, obconic at base, 1.9-2.5 mm. in diameter; petals whitish (drying yellowish), not expanding, obscurely graduated, often shorter than the calyx-teeth; banner little recurved, oval, subentire, 4.3—7.2 mm. 2.5-3.8 mm. wide; wings 4.2-6.5 mm. long, the claws 1.5-2.9 mm., the blades 3-4.5 mm. long, 1-2 mm. wide; keel 4.2-6 mm. long, the claws 1.8-3.2 mm., the blades 4.2-3.6 mm. long, 1.3-1.9 mm. wide, the whole often carried up on the style of the forming fruit; anthers mostly smaller, 0.3-0.45 (0.5) mm. long; pod (alike on both types of inflorescence) ascending or loosely spreading (humistrate), ovoid-acuminate, ovoid-, lance-, or narrowly oblong-ellipsoid, 1.2—3.7 cm. long, 5—8 mm. in diameter, straight, lunately incurved, or rarely a little decurved, rounded or narrowly to broadly cuneate at base, either abruptly contracted or gradually tapering distally into a triangular- or lance-acuminate, somewhat compressed beak, dorsally flattened or openly depressed-sulcate (when obscurely trigonous) in the lower ?-?, carinate ventrally by the prominent, thick suture, the dorsal suture also prominent but thinner and commonly undulate, the green, somewhat fleshy, strigulose or villosulous valves becoming stramineous or purplish-brown, leathery or stiffly papery, cross- reticulate and often wrinkled lengthwise, not inflexed; seeds ocher- or pale brown, often purple-speckled, irregularly pitted, dull or lustrous, 1.5-2.5 mm. long.— Collections: 133 (xxii); representative material cited in the discussion.
Prairies, barren hilltops, gullied bluffs, and gravelly outcrops on plains and hillsides, sometimes (northward) on sandy or pebbly lake shores, usually in light and porous or gravelly clay soils, sometimes in black loam, limestone silt, shaley gravels, or almost pure gypsum, mostly 700—6000 feet, but ascending westward along streams into the valleys of the southern Rocky Mountains up to 7800 feet, common on the prairies of central Texas and Oklahoma, north, becoming less frequent (but because of the prevalence of cleistogamy and inconspicuous fruits probably much overlooked), to Minnesota, southern Manitoba, and the Saskatchewan and Red Deer Valleys in Alberta, west across the Continental Divide to the lower Fraser and Kootenai Valleys in southern British Columbia, in Montana to the Bighole River in Beaverhead County, in Wyoming to the Big Horn and
Wind Rivers, in Colorado to the foothills of the Front Range and up the Arkansas Valley into Chaffee County, and in New Mexico to the upper Rio Grande; greatly isolated, perhaps introduced, in southern Coahuila at 6300 feet (Ripley & Barneby 13,267), this record not mapped.—Map No. 139.—April to July.
Astragalus lotiflorus (with flowers of Lotus) Hook., Fl. Bor.-Amer. 1: 152. 1831.— "About Carlton-House on the Saskatchewan, Drummond."—Holotypus, labeled "North America," K!—Phaca lotiflora (Hook.) T. & G., Fl. N. Amer. 1: 349. 1838. Astragalus lotiflorus fma. pedunculosus (with peduncled racemes) Gray in Proc. Amer. Acad. 6: 209. 1864. Tragacantha lotiflora (Hook.) O. Kze., Rev. Gen. 946. 1891. Cystopora lotiflora (Hook.) Lunell in Amer. Midl. Nat. 4: 428. 1916. Batidophaca lotiflora (Hook.) Rydb. in N. Amer. Fl. 24 : 321. 1929.
Phaca cretacea (of chalky soil) Buckl. in Proc. Philad. Acad. 1861: 452. 1861.—"Cretaceous rocks, Northern Texas."—Holotypus, collected in May, 1861, by S. B. Buckley, PH!— Batidophaca cretacea (Buckl.) Rydb. in N. Amer. Fl. 24: 322. 1929. A. lotiflorus var. cretaceus (Buckl.) Gates in Trans. Kans. Acad. Sci. 42: 137. 1940.
Astragalus lotiflorus fma. brachypus (shortly peduncled) Gray in Proc. Amer. Acad. 6: 209. 1864.—"Both forms are represented in Hall and Harbour’s collection, no. 131."—Holotypus, GH! isotypus, G!—A. elatiocarpus (from elatior, taller and carpos, pod, the peduncles supposed to elongate after the fruit is set) Sheld. in Minn. Bot. Stud. 1: 20. 1894, based in large part on the preceding. A. ammolotus (sand lotus) Greene in Erythea 3: 76. 1895, a pedantic and illegitimate substitute. A. lotiflorus var. elatiocarpus (Sheld.) Rydb. in Mem. N. Y. Bot. Gard. 1 (Fl. Mont.): 244. 1900. Phaca elatiocarpa (Sheld.) Rydb. in Bull. Torr. Club 32: 665. 1905. Cystopora elatiocarpa (Sheld.) Lunell in Amer. Midl. Nat. 4: 428. 1916.
Astragalus Reverchoni (Julien Reverchon, 1834—1905, collected extensively in Texas from 1856 onward) Gray in Proc. Amer. Acad. 19: 74. 1883.—" ... in coll. N. American Plants, distributed by A. H. Curtiss, no. 601.A, supplied by J. Reverchon, from the central parts of Texas..."—Holotypus, Reverchon 8, from n.-w. Texas, collected in 1880, GH! isotypi, 2 sheets, NY!—Phaca Reverchoni (Gray) Rydb. ap. Small, Fl. S. E. U. S. 1132. 1903.
Astragalus lotiflorus var. nebraskcensis (of Nebraska) Bates in Amer. Nat. 29 : 670. 1895.—"On June 25, 1892, in the outskirts of Long Pine, Brown Co., Nebraska ... Specimens ... in the herbarium of the Botanical Survey of Nebraska, University of Michigan, and Columbia College."—Lectotypus, labeled by Bates as "the original find," NY! isotypus, US!—A. nebraskensis (Bates) Bates in Torreya 5: 216. 1905. Batidophaca nebraskensis (Bates) Rydb. in N. Amer. Fl. 24 : 322. 1929.
Astragalus Batesii (John Mallory Bates, 1846—1930, from 1902 Episcopal minister at Red Cloud, Nebraska, ornithologist and botanist) A. Nels. in Bot. Gaz. 54: 150. 1912.—"[Bates] no. 5501, Red Cloud, Neb., May 17 and May 23, 1911."—Cotypi, mounted together, RM!
The lotus milk-vetch is a perplexingly variable and polymorphic species. On account of the floral dimorphism, the plants within a colony have actually or potentially a dual aspect; and the populations differ from one to the next in length, density, and orientation of the pubescence, in size and color of the developed flower, and in length and outline of the fruit. Rydberg (1929, p. 315) maintained three species defined mainly in terms of pubescence and pod characters, but these are far from being clear-cut entities and their ranges all appear to converge and overlap in Nebraska and adjoining states. Jones (1923, p. 177) described some of the variants of A. lotiflorus, but concluded that they were completely confluent and beyond reach of taxonomic delimitation. However it is possible to make out a rough correlation between a northern and western dispersal, long-pilose or occasionally villous herbage, and a relatively short-beaked, ovoid, villosulous pod (A. lotiflorus, sens, str.); representative: J. Macoun 10,192 (ND); Moodie 250 (NY); Hitchcock & Muhlick 11,802 (NY, RSA, WS); Bush 199 (NY); Rydberg 74 (NY); Osterhout 4327 (NY); Ripley & Barneby 8316 (CAS, RSA); Jones 43 (NY, POM); Warnock 5996 (SRSC), showing a range from Alberta to western Texas; and to contrast this with a southeastern pattern of dispersal, a vesture of shorter, subappressed hairs, and a long, distally acuminate, strigulose fruit (var. Reverchoni); representative: Cory 55,605 (SMU, WS); E. Hall 150 (NY); Reverchon (from Sweetwater, Texas) in 1882 (NY, SMU); Shinners 9712, 12,180, 12,380 (SMU); Waterfall 5948 (OKLA); G. W. Stevens 145 (OKLA, NY); Ripley & Barneby 82S2 (CAS, RSA), 10,567 (RSA); Eggleston 20,095 (NY), the characteristic Texan and Oklahoman plant. In the first category cleistogamy is very common, developed flowers being almost unknown on the higher prairies and in the Rocky Mountain foothills, whereas in Texas and Oklahoma the pedunculate inflorescence is the normal one and cleistogamy occurs only rarely, mostly in seedling plants. The following key, prepared during the preliminary studies of Lotiflori, presents an idealized view of the possibilities of recognizing two named varieties in the species:
1. Hairs of the herbage mostly straight, appressed or ascending at a narrow angle, up to 0.5—1.1 (a few sometimes to 1.5) mm. long; pod mostly over 2 cm. and up to 3.7 cm. long, commonly acuminate distally, strigulose, often silky-canescent in youth; centr. Texas and Oklahoma n.-ward (becoming rarer) to Minnesota and South Dakota = var. Reverchoni (including Phaca cretacea and A. Batesii).
1. Hairs of the herbage either straight and loosely spreading to widely ascending, or spreading and sinuous, mostly up to 1.2-2.2 (2.5) mm. long; pod 1.2-2.7 cm. long, more shortly beaked, villosulous or villous-hirsute; s. Canada to trans-Pecos Texas, w. to Montana, Wyoming, and Colorado, passing e. and s.e. into the last = var. lotiflorus (including A. elatiocarpus, p. p., and var. nebraskensis, a variant with particularly long and loose pubescence).
Near the northwest and southeast limits of the range of A. lotiflorus, all plants encountered are likely to satisfy the conditions set up in the key, but unfortunately this is not so elsewhere. In the upper Mississippi Valley and adjoining states the majority of plants are intermediate in either one or all differential characters. Furthermore there are instances of one feature appearing geographically isolated deep in the wrong territory, e.g., A. & G. Heller 3768 (ND, NY) from Rio Arriba County, New Mexico, which combines in the cleistogamous phase the type of vesture to be expected at the west limit of the species with the long fruit characteristic of var. Reverchoni on the low southeastern prairies. Occasionally, also, specimens technically referable to both varieties have been collected in the same place and distributed under one number (e. g., G. W. Stevens 3050, OKLA). The evidence before me justifies Jones’s treatment of A. lotiflorus as a polymorphic but, in practice, indivisible entity.
From the much greater frequency of normally developed racemes in Texas and Oklahoma it seems fair to assume that the so-called var. Reverchoni represents the primitive state of the lotus milk-vetch. It is a modest, pretty astragalus, easily recognized by the combination of dolabriform vesture with a tufted habit of growth, free stipules, and campanulate calyx-tube. In its area of dispersal A. missouriensis alone is like it in pubescence and habit, differing in the much larger, proportionately narrower flower. Southward, the petals vary from whitish through increasingly vivid shades of lavender into a lively purple, the depth of color seemingly becoming accentuated on calcareous or gypseous soils. The common northern phase of the species is cleistogamous and forms an inconspicuous tuft of pinnate foliage, to the casual glance suggesting some seedling astragalus too young to flower; and it is only close scrutiny which reveals the tiny flowers in the leaf-axils or the subradical clusters of pods huddled close to the ground in the shelter of the leaves. In the rarer chasmogamous individuals which occur northward, the petals vary from whitish to cream-color or pale lilac. Exceptionally both sorts of raceme arise on one plant, as happens also southward, and in them the developed peduncles appear in spring, followed by the cleistogamous type in summer. The causes of cleistogamy are not known, but a long photoperiod may be required. In the southern states, where cleistogamy is rare, established plants reach anthesis soon after the spring equinox; these are followed sometimes by cleistogamous seedlings which flower from seeds germinating the same year. At greater elevations and more northern latitudes, anthesis is delayed until the day is appreciably longer. Occasionally there occurs a type of inflorescence intermediate between the fully cleistogamous and fully developed one; it consists of a short peduncle and about five small pallid, perfectly formed flowers. When Sheldon described A. elatiocarpus, he believed that the subobsolete peduncle already described by Gray (as fma. brachypus) continued to develop after pollination of the flowers and eventually lifted the ripening pod up to the level of the leaves. Nothing of the sort is known to occur, and Sheldon’s term "elatiocarpic" was a needless invention. Although A. elatiocarpus was based in part on fma. brachypus, the plant actually observed by Sheldon (No. 5154, MINN, WIS, from Ottertail County, Minnesota) is an instance of a vernal chasmogamous flowering giving way on the same plant to a later cleistogamous one, both types of inflorescence persisting on a single plant.
The Mexican record of A. lotiflorus is based on a single plant of the cleistogamous type collected on a weedy roadside in Carneros Pass south of Saltillo, Coahuila. More collections are required before the species can be considered truly native or even truly established south of the border. For the present A. lotiflorus has been omitted from the key to the Mexican species, and the Mexican record left unmapped.