132c.  Dalea versicolor Zuccarini var. calcarata (Gentry) Barneby

(Plate CXIX)

Suffruticose becoming shrubby, the stems 1 or few together, irregularly branched in age, (4) 5-15 dm tall, at first openly paniculate distally and bearing a few pedunculate spikes, later flowering from axillary short-shoots; stipules 0.7-3 mm long; foliage glabrous or pilosulous, when the latter greenish, the leaflets pubescent both sides or only beneath; leaflets of primary leaves 8-17 ("19") pairs, usually not strongly carinate beneath; spikes 1.3-1.6 cm diam, the axis 0.5-5 (6) cm long; calyx (sessile or minutely pedicellate) 6.5-8 mm, the tube 2.5-3.4 mm, the dorsal tooth 3.3-5 mm long; epistemonous petals lilac-pink to rose-purple.— Collections: 12 (o).

Open grassy slopes in the oak or oak-pine belt, 1300-2220 m (± 4330-7400 ft), w. slope of Sierra Madre Occidental from s. Sonora (Sierra de Alamos) and the canyons tributary to Rio Fuerte in s.-w. Chihuahua s. through Sinaloa and adjoining Durango just to the Tropic line. — Flowering November to April. —Material: Sonora. Alamos: Palmer 282 (NY). Chihuahua. Urique: Caddell 23 (OKLA). Sinaloa. Badiraguato: Gentry 5811 (ARIZ, MICH, NY), 7180 (NY), 7291 (F, NY, UC). San Ignacio: Ortega 34 (MEXU). Durango. Santiago Papasquiaro: Palmer 138 (C, F, NY, UC). El Salto: Gilly 11,550 (MICH, RENNER); McVaugh 23,604 (MICH); Breedlove 18,910 (NY).

Dalea versicolor Zucc. subsp. versicolor var. calcarata (Gentry) Barneby, stat. nov., based on D. wislizeni subsp. calcarata (spurred, of the calyx-teeth) Gentry, Madrono 10: 243, pl. 21, 22, fig. 3 (map). 1950.— "Los Pucheros, Sierra Surotato, northern Sinaloa, Mexico...March 24, 1945, Gentry 7182.." — Holotypus, not examined, DS; isotypi, F, MICH, NY, UC, US!

Dalea surotatensis (of Sierra Surutato) Gentry, Madrono 10: 234, pl. 17. 1950.-"Above La Jolla, Sierra Surotato, Sinaloa, Mexico...March 17-24, 1945, Gentry 7291a..." — Holotypus, not examined, DS; isotypi, F, MICH, NY, US!

The var. calcarata is weakly distinguished from the vicariant var. sessilis by the more numerous leaflets of primary leaves. Even when it was described, Gentry admitted that there were ambiguous intermediate forms in southern Sonora. The concept of var. calcarata, based in the first instance on a drought-inhibited plant glabrous to the spikes, is here expanded to accomodate D. surotatensis, described from densely pilosulous plants collected at the stage, paradoxically in March, of maximum growth activity usually associated with autumnal flowering. This expansion might be thought, a priori, excessively liberal until it is reexamined in context of D. versicolor sens, lat., and in light of information provided by other populations of the species present in central and southern Sinaloa and adjoining Durango. A plant collected long ago by Edward Palmer at San Ramon, on headwaters of Rio de los Remedios(No. 138, cited supra), which had been referred by Rydberg (in herb.) to D. saffordii and mentioned by Gentry (1950, p. 242) as anomalous subsp. wislizeni, is believed to represent var. calcarata with glabrous leaflets but still pubescent leaf-stalks. A little further south in Durango McVaugh and Breedlove recently collected another form of the same with thinly pubescent foliage. Pubescence cannot be considered taxonomically reliable in this group. Gentry considered D. surotatensis akin to D. pulchra rather than to any form of what is here called D. versicolor, his opinion apparently relying heavily on glandless wing-petals as a differential character. These glands (present on keel and banner in D. surotatensis) are lacking in numerous examples of D. versicolor sens, lat., among others in Palmer 138 (NY) just mentioned.

It is certainly remarkable that the detached Sierra Surutato should harbor both glabrous and villosulous populations of one entity, but the same thing occurs in central Oaxaca in the case of var. versicolor. The type-series of D. surotatensis shows mainly terminal long- pedunculate flower-spikes such as are expected only in the fall, but Gentry collected (no. 7180) in the same week and same vicinity the drought-inhibited phase in which the virgate primary stem has lost large leaves and is clothed in short-shoots bearing short but not especially dense spikes. It remains to be noted that D. surotatensis does differ in one other point from other phenotypes of var. calcarata: the leaflets are charged dorsally with smaller and more numerous glands of a type that is common in var. versicolor and occurs, very occasionally, in var. sessilis also. But there seems to be no perfectly reliable differential character in any form of D. versicolor.