Chamaecrista lineata var. lineata

Shrubs mostly 5-12 dm, in exposed places freely branched and ultimately of bushy or globose outline, in thickets sometimes of spindly stature and up to 1.5-2 m, often flowering the first season as erect, few-branched or simply virgate, weakly suffrutescent herbs, the main trunk reaching 2 cm diam but mostly less, the annotinous leafy branchlets and lf-stalks strigulose or pilosulous with all straight appressed, all straight ascending-spreading, rarely all incumbent, whitish hairs to 0.15—0.4 mm, exceptionally mixed with few ascending setules up to 0.4— 0.7 mm, the foliage usually bicolored, the firmly chartaceous lfts above dull olivaceous brunnescent, beneath commonly but not always glaucescent, strigulose- pilosulous like the stems on both faces, or on the lower only, or less often merely ciliolate, only exceptionally quite glabrous; stipules (l-)l.5-5(-6.5) x 0.6-1.2 (-1.4) mm; major lvs to (1.2-) 1.5—5(—5.5) cm; gland (0.25—)0.35—1.2(-l .8) mm diam, usually sessile or almost so and either sunk within the ventral groove of petiole or resting at level of the wing-margins, locally (Jamaica; e. Cuba; n.-w. Haiti) shortly and stoutly stipitate when in profile squatly tack-shaped and up to 0.40.7 mm tall, exceptionally a trifle longer than diam of head; rarely similar glands on 1-3 segments of rachis; lfts of major lvs (1—)2—7, locally in Bahamas (Great Bahama, New Providence, Andros), Jamaica, e. Cuba (Oriente) up to 8-11 pairs, very variable in size and outline but when relatively many (5+ pairs) mostly oblong-oblanceolate or -elliptic and when fewer (1-5 pairs) obovate, obovate- elliptic or -cuneate, obtuse mucronulate, obscurely acuminulate mucronulate, or widely open-emarginate, the larger up to (7.5-) 10-21 (-24) x (2.5—)3—8.5(—10) mm, the secondary major venules from abaxial side of midrib (3-)4-9(-10), with or without finer intercalary ones, the venulation always prominulous beneath but variable in strength, usually finer so above; peduncles adnate through 0-3.5 mm; sepals up to (6—)7—10(—11) mm; long petals (8—)9—16 mm; ovules (8-)9-16; pod 25-50 x 3.4-5 mm.—Collections: 155.

Dunes, sand flats behind barrier beach, soil pockets in limestone (coral) pavement, limestone rock ledges, and extending inland in coppice thickets, savanna and palm-savanna, becoming locally dominant and sometimes especially vigorous and large-leaved in waste ground or along highways, over most of its range maritime and submaritime below 5 m, ascending inland in Jamaica and eastern Cuba (Sa. de Nipe) to ±700 m, in centr. and e. Cuba locally abundant on ferruginous and serpentine outcrops, widespread through the Bahamas s.-ward from Bimini, Grand Bahama and Abaco to the Caicos Is. and n.-w. corner of Hispaniola (Ile de la Tortue, St. Louis du Nord, Port Margot); s.-w., centr. and e. Cuba (Isla de Pinos; Matanzas, Las Villas, Camaguey, Oriente, but locally replaced in Sa. Maestra and vicinity by vars. brachyloba and brevipila); w. and n. coastal Jamaica; Grand Cayman.—Fl. interruptedly through the year, barren only during drought.

The length of our varietal description reflects the great latitude of variation encountered in var. lineata, variation that independently involves growth-habit, length and diameter of petiolar gland, vesture of stem and foliage, and number, outline and amplitude of leaflets. The patterns of variation seem to arise both from genetic differences, little variable within a population or within a given geographic limit, and from external influences of microenvironment and season; but the interplay of these two sets of forces is cryptic and will remain so until recourse can be had to controlled experimental plantings. Nevertheless our analysis of the variable features has brought out some interesting facts:

a) It is a myth, propagated by Britton & Rose’s key to Chamaecrista ser. Lineatae (1930, p. 271) that the foliage of Ch. lineata, even as defined in the form of specimens at NY annotated in Britton’s hand, is "glabrous or nearly so." In fact truly glabrous leaflets are exceptional, unknown from Jamaica (the type- region), whereas leaflets strigulose beneath or on both faces are the rule, and leaflets pilosulous with patent hairs on one of both faces occur at random points in the Bahama Chain (Grand Bahama, Correll 44763; Exuma, Correll 40725; San Salvador, Gillis 8818; Acklin, Brace 4290, all NY), but not, so far as known, beyond.

b) The petiolar gland is in the Bahamas uniformly sessile or almost so, commonly sunk in the ventral groove of the petiole or seeming in profile view to rest snugly on its narrow ventral flanges; but in either situation varying greatly in size, and in color from red to brown, yellowish, or green. In Jamaica, both on the seashore and inland on the Santa Cruz Mountains, as also in eastern but not in central Cuba or Isla de Pinos, and in coastal Haiti (Nord-Ouest) the gland is elevated on a distinct stalk, and becomes essentially like that of var. brachyloba, even while still associated with the distinctive glaucous foliage of Bahaman var. lineata.

c) Leaflet number, described here in terms of the highest number found in given plant’s largest leaves, appears to be relatively stable within major populations and even over considerable segments of the range of var. lineata; but a survey of the whole range shows random inconsistencies and distantly allopatric duplications of phenetically identical types. In the Bahamas northward from New Providence a maximum of 6-8 pairs prevails; southward a maximum of 2-5 pairs, the numbers commonest in central Cuba (Matanzas, Las Villas, Camaguey), and never exceeded in Isla de Pinos, Grand Cayman, or Haiti. But Jamaica and the Caicos yield records of 3-7 and 5-7 pairs respectively, eastern Cuba, both maritime and interior, 3-8 pairs. Extremes of 9-11 pairs, which must be considered quite exceptional, occur on New Providence (Curtiss 180, A, G, GH, NY, US; Britton & Brace 397, NY) and on Sa. de Nipe in eastern Cuba (Shafer 8414, NY). Because number and shape of leaflets are to some degree correlated, a loss in number being compensated for by increase in amplitude, the forms with 1-3- jugate and 5-8-jugate leaves appear strikingly different, but are connected by forms, just as numerous and without clear geographic grouping, of intermediate number. We find no relation whatever between leaflet number and pubescence.

Of the synonyms listed above, Cassia cuneata remains somewhat doubtful, due to the fragmentary nature and unknown origin of the holotypus. The name has never hitherto been associated with anything more substantial than the type- specimen itself, and was supposed by Bentham to differ from C. lineata in having 4-6, not 2-4 pairs of leaflets, a character now known to be valueless and even at the time irrelevant, for Swartz’s original C. lineata from Jamaica had five pairs. The typus of Ch. clarensis represents the minor variant of C. lineata in which the 2-3-jugate leaflets are puberulent on the upper as well as the lower face. That of C. niqueroensis represents the extreme reduction of the leaves to a single or (in the isotypus at NY) in some few leaves 2 pairs of leaflets. Approximate to- potypes (León 16335, GH, US) have many, or (US) mostly bijugate leaflets and are essentially identical with the races of var. lineata found distantly allopatric on the coast of Haiti. This extreme form was collected first somewhere in Cuba by Ramon de la Sagra and annotated by Richard (P) as an undescribed species; but his manuscript name seems to have remained no more than that.