Astragalus hartwegii Benth.

Caulescent perennial, with a taproot and root-crown or shortly forking caudex at or just below soil-level, varying from quite slender to comparatively stout and coarse, strigulose with appressed or subappressed hairs up to 0.25-0.5 mm. long, the herbage green or cinereous, the leaflets either glabrous or thinly pubescent above; stems seemingly diffuse and incurved-ascending, 1-3 (4) dm. long, slender at base becoming thicker distally, simple, or branched or spurred near the base; stipules 1-4 mm. long, subdimorphic, the lowest ones small, scarious, pallid, fully amplexicaul but free, the upper ones triangular or lanceolate, herbaceous or firmly papery, semiamplexicaul, the blades often reflexed; leaves (2) 3-9 (15) cm. long, all but the lowest subsessile, with 15-25 (29) oblong or oblong-elliptic to narrowly oblanceolate or almost linear, obtuse or subacute (exceptionally, in some lower leaves, obovate and emarginate), flat or loosely folded leaflets 3-18 (21) mm. long; peduncles incurved-ascending, (4) 6-16 (20) cm. long; racemes densely or at length loosely (15) 25-55-flowered, the flowers early reflexed, either secund or loosely retrorse-imbricated, the axis more or less elongating, (2) 4-14 cm. long in fruit; bracts membranous, pallid, ovate or lanceolate, 1-4 (5) mm. long, early deflexed; pedicels arched out- and downward, at anthesis 0.2-0.8 mm., in fruit thickened, persistent, 0.6-1.3 mm. long; bracteoles 0; calyx 3.2—4.8 mm. long, loosely strigulose with white or largely fuscous hairs, the strongly oblique disc 0.6-0.8 mm. deep, the campanulate tube 2.2-3.4 mm. long, 1.7-2.5 mm. in diameter, obliquely truncate or commonly a little gibbous dorsally at base, the subulate or triangular-subulate teeth 1—1.4 mm. long; petals (at least the banner) purplish, drying pallid or bluish; banner recurved through 45-50°, rhombic- ovate-cuneate, notched, 7-9 mm. long, 3.8-5.2 mm. wide; wings (0.3 mm. longer to 0.4 mm. shorter) 6.6—8.6 mm. long, the claws 2.6—3 mm., the lanceolate, obtuse blades 4.5-6.3 mm. long, 1.2-1.6 mm. wide, gently incurved in the distal half; keel 5.4—6.3 mm. long, the claws 2.6—2.9 mm., the half-obovate or obliquely obovate blades 3.1-4.2 mm. long, 1.9-2.1 mm. wide, incurved through ± 90° to the rounded apex; anthers 0.4-0.5 mm. long; pod deflexed, sessile, deciduous (sometimes tardily) from the conical receptacle, narrowly lanceolate or linear- oblong in profile, gently incurved or almost straight, 1—1.5 cm. long, 1.8—2.4 mm. in diameter, rounded at base, contracted into a short subulate cusp at apex, trigonously compressed, carinate ventrally by the suture, openly sulcate dorsally, the thin, green, strigulose valves becoming papery, stramineous or ultimately blackish, delicately cross-reticulate, inflexed as a complete or almost complete septum 0.8—1.2 mm wide; dehiscence apical and downward the length of the ventral suture, and ultimately through the septum, the halves of the pod tardily separating; ovules (12?) 14-18; seeds brown or olive-brown, heavily purple-speckled, often prismatically compressed, smooth and sublustrous, 1.2-1.5 mm. long.—Collections: 14 (ii); representative: Palmer 290 (GH, NY, US); Purpus 5195 (GH); Parry & Palmer 168 (GH); Palmer 65 & 81 (GH, NY); Rose & Hay 5286 (US); Waterfall 15,520 (NY).

Plains and hillsides, ranging from arid grassland at 3000 feet into open oak woodland and pine forest as high as 8200 feet, not uncommon in northcentral Mexico, from near Ciudad Durango south and southeast to Guanajuato, and through San Luis Potosi to Hidalgo.—Map No. 155.—June to October.

Astragalus Hartwegi (Karl Theodor Hartweg, 1812-1871, collected extensively in California and Mexico) Bth., Pl. Hartw. 10. 1839.—No locality given, but the plant found in Hartweg's collection of 1837, gathered on the road between Mexico City and Zacatecas.—Holotypus, Hartweg 53, labeled "Aguas Calientes," K (herb. Bth.)! isotypi, BM, G, GH, OXF, P!—Tragacantha Hartwegi (Bth.) O. Kze., Rev. Gen. 945. 1891 ("Hartwegii"). Hamosa Hartwegi (Bth.) Rydb. in Bull. Torr. Club 54: 333. 1927 ("Hartwegii").

Hamosa hidalgensis (of Hidalgo) Rydb. in Bull. Torr. Club 54: 333. 1927.—"Type Pringle 9720 (herb. N. Y. Bot. Gard.) was collected at Dublan, Hidalgo, September 19, 1902 ..."— Holotypus, NY! isotypi, GH, K, US!

The Hartweg milk-vetch, one of several Mexican Astragali with many tiny, reflexed flowers arranged in long, rat’s-tail spikes, is characterized by rather minute technical features and has been confused repeatedly with the closely related A. vaccarum and with A. (Strigulosi) micranthus var. micranthus. These three species have not been clearly separated since 1881, when Hemsley in Biologia Centrali-Americana cited as A. Hartwegi material of A. vaccarum from Sonora (Schott) and of A. micranthus from near Mexico City (Bourgeau 337, 483). The dispersal northward "probably to Chihuahua" attributed to A. Hartwegi by Jones (1923, p. 278) is certainly based on mistaken identities, chiefly involving A. vaccarum. Finally Rydberg’s Hamosa Hartwegi (l.c.) can be analysed into the same three components, all material cited from Chihuahua, Sonora, and Coahuila representing A. vaccarum, whereas the Bourgeau and Arsène collections from Hidalgo and the Federal District are A. micranthus. The differential characters of A. vaccarum, already brought out in the sectional key and discussed under the following species, need not detain us here. Those of A. micranthus are easily observed only in complete and carefully collected specimens which show the connate lower stipules of sect. Strigulosi and, if possible, the mature pod. The fruit of var. micranthus is variable in size and curvature, but it resembles that of A. Hartwegi really closely only in a small proportion of the specimens examined, those representing the form that has been called A. saltonis Jones. But even in these the pod remains attached to the receptacle until dehiscence and, as compared with that of genuine A. Hartwegi, is of perceptibly thinner texture and less sharply trigonous compression, having blunter lateral angles and more convexly rounded lateral faces. In most forms of A. micranthus the ovules are 4-7 pairs, whereas 7-9 pairs are commonest in A. Hartwegi; this character is indecisive, but often useful. While two perfectly distinct species unquestionably exist, I must admit that they are often astonishingly alike in growth-habit and even in fine detail of the floral structure. The similarities are of a sort to challenge the validity of a phylogenetic system which contrives to refer the two species to different sections and implies a widely separated origin for each. It is an accepted fact that pairs of organisms may mimic each other by means of evolutionary convergence, but it is difficult to prove convergence in members of the same genus, where a similar facies is more likely than not to denote a genuinely close relationship and near common origin. However there can be no reasonable doubt that A. Hartwegi forms with A. vaccarum and A. Goldmani a natural group of piptoloboid species allied to sect. Leptocarpi; and it seems equally sure that A. micranthus is a true member or the emmenoloboid sect. Strigulosi, with close relations in A. Purpusi and A. pueblae. If the deep dichotomy assumed to exist between the emmenoloboid and piptoloboid phyla in Astragalus corresponds with a historic reality, A. micranthus and A. Hartwegi ought not to be near allies but simply a convergent pair. Possibly, however, the skein is not so simply woven; it is suspected that anastomosing strands crossed over in the past from one major gene-stream into another, giving rise to bridges and confusing passages between the main lines of heredity in the genus, which have elsewhere remained, broadly speaking, intact and pure. If it is possible to conceive of introgression between the forebears of the small-flowered Strigulosi and the Micranthi, characterized on the one hand by buried root-crown, connate stipules, and persistent pod, and on the other by superficial crown, free stipules, and deciduous fruit, we should expect to find various combinations of these characters handed down to contemporary forms. It is perhaps no coincidence that two astragali flourishing today on the Mexican Plateau fulfil just these conditions, and at the same time have tiny, reflexed flowers and small, bilocular fruits quite similar to those of A. Hartwegi and A. micranthus. These two are A. oxyrrhynchus, which has all characters of the Micranthi except the buried root-crown, a feature very rarely combined with free stipules; and A. hypoleucus, closely resembling the last-mentioned except for the sheathing stipules, and thus suggesting a member of sect. Strigulosi with disjointing pod. I cannot help suspecting that all share in a common heritage. The situation is perhaps analogous with that described in connection with sects. Trichopodi and Oxyphysi in California.

Variation in A. Hartwegi is of a superficial nature involving stature and vigor, width of leaflets, and size or curvature (or both at once) of the pod, which may be straight or gently incurved. The vesture of the herbage varies in dispersal over the leaflet surface, and in color on the calyx and pedicel. A robust plant with relatively small flowers and dark-hairy inflorescence was singled out by Rydberg to typify his Hamosa hidalgensis, but it is only one of several known minor variants. The only other collection (Rose 2757 from Plateado, Zacatecas) associated by Rydberg with H. hidalgensis represents A. esperanzae. A topotypus of H. hidalgensis with paler vesture (Rose & Hay 5286, US) was annotated by Rydberg himself as H. Hartwegi.

The Hartweg milk-vetch was collected first late in the XVIII century by Sessé & Mocino (OXF).