Mapping is provisional and incomplete. Var. calvescens is excluded; see notes on characters under that name. There has been some combination of these two taxa in older treatments (D, B, F); and there may be some intergradation. There has also been confusion with "ciliosa" of Sm, who included some midwestern humilis with R. ciliosa Pursh, which in its strict sense is restricted to the Coastal Plain east of the Mississippi Rv. (W). R. humilis, as treated here, is still a variable species, mostly mid-western, locally abundant in the Black Belt of Miss. and Ala., and probably scattered across western regions of Tenn. and Ky. Records from Ky. are mostly somewhat obscure (M) but they include colls. of BA from CHRI (TENN), B from EDMO (US), McFarland from HART (TENN), Palmer (as the large-leaved "var. frondosa") from LOGA (MO) and JC from WARR (Petros-Browning Rd). BA's report of the large-flowered "var. longiflora" is based on a coll. of calvescens (Wharton #8793 at EKY). Also, humilis has been cultivated in some wildflower gardens of Lexington for several years, and continued observation confirms their distinction from calvescens; they become mostly cleistogamous later in the growing season (B. Popkin, pers. comm.). Such plants are widely distributed, usually under the name humilis (e.g., see website of Missouri Botanical Garden). [Sm treated ciliosa as a widespread species, with subsessile or sessile leaves like humilis, upright habit like caroliniensis (up to 5-7 dm tall), dense spreading pubescence (typically more hairy than caroliniensis or calvescens), and often larger flowers (up to 5-7 cm long). Ward (2007) treated it as R. caroliniensis ssp. ciliosa (Pursh) R.W. Long, a "dimorphic or heteromorphic" taxon. Most recently, W has treated ciliosa as a distinct species of the sandhills from Miss. to N.Car., with mostly basal leaves, and apparently closer to caroliniensis than humilis.]