GILIA FAMILY 



177 



Creek, Lassen Co., Nutting; Big Valley, near Lookout, Modoc Co., Baker 4' Nutting. North 

 Coast Ranges: Yollo Belly Mts., T. Brandegee; Humboldt Co., Tracy 8864 (Brannan Mt.), 7679 

 (Horse Mt.), 3365 (betw. Three Creeks and Willow Creek), 10,374 (Trinity Summit); Union 

 Creek, Salmon Mts., Hall 6595 ; Poker Flat, w. Siskiyou Mts., R. Van Deventer 267 ; Klamathon, 

 Siskiyou Co., Copeland 3560. 



Eefs. — GiLiA CAPiLLARis Kell., Proc. Cal. Acad. 5:46 (1873), type loc. Cisco, Placer Co., 

 Sierra Nevada, Kellogg ; Jepson, Man. 799 (1925). G. leptalea subsp. capillaris Brand; Engler, 

 Pflzr. 42=0:98 (1907). G. Unearifolia Howell, Fl. Nw. Am. 461 (1903), type from s. Ore. G. sub- 

 alpiiia Greene; Brand, I.e., type loc. Donner Lake, Heller 7042. 



5. G. tenuiflora Benth. (Fig. 381.) Stem erect, paniculately branching above, 

 7 to 20 inches high, glabrous or glandular-puberulent above ; leaves mostly basal, 

 1 to 2 inches long, glabrous or puberulent, the blades once or twice piunately divided 

 into small segments % to 1 line long; flowers on 

 short or long pedicels in a loose panicle or cyme, 

 solitary or clustered; calyx 1 to 2V2 lines long, 

 its intervals scarious, its acute teeth i/4 to V2 as 

 long as the tube; corolla fmmelform, purple with 

 dark purple throat, 3 to 8 lines long, its tube usu- 

 ally not exceeding the calyx (sometimes 2 to 3 

 times as long) , gradually enlarged into the throat 

 (in type) or sometimes abruptly expanded, the 

 throat usually much longer than the tube, the 

 limb 3 to 5 lines broad; stamens unequal, in- 

 cluded or slightly exserted, the filaments equally 

 inserted below the corolla-sinuses ; capsule many- 

 seeded, equaling or exceeding the calyx; seed- 

 coats developing mucilage when wetted. 



Sandy flats and hill slopes, 50 to 2200 feet: 

 South Coast Ranges from Santa Clara Co. to San 

 Luis Obispo Co. and Santa Rosa Isl. Apr.-June. 



Locs. — Coyote sta., Santa Clara Co., Jepson 15,086; 

 Mission Soledad, Monterey Co., Jepson 8434a ; upper San 

 Antonio River, Monterey Co., Jepson 1657; betw. Brad- 

 ley and San Miguel, Salinas Valley, Wieslander ; Paso 

 Robles, Barher ; Cholame, e. San Luis Obispo Co., Jepson 

 16,189; Elkhorn Scarp, se. San Luis Obispo Co., Jepson 

 16,231 ; Old Ranch Caiion, Santa Rosa Isl., Hoffm-ann. 



Geog. note. — The type of Gilia tenuiflora Benth. is 

 a Douglas plant from California and therefore came from 

 the Coast Ranges, most likely from Monterey Co. or San 

 Luis Obispo Co., since a collection made on the Mission 

 Trail at Paso Eobles {Barter) matches very closely the type preserved in the Royal Botanic Gar- 

 dens at Kew. Even within the somewhat narrow limits of the Salinas Valley of Monterey and 

 San Luis Obispo counties and its bordering hills, Gilia tenuiflora displays a certain variability 

 as to herbage and relative length of corolla-tube and calyx that is prophetic of the great variations 

 found in the species southward and in the transmontane deserts. In those regions it breaks into 

 a large number of diverse forms exhibiting corolla variables and leaf variables. So great is the 

 variability in size and shape of corolla and in leaf structure that the extremes, often published as 

 species, might well be regarded as valid species save for the truly overwhelming number of inter- 

 grade forms representing not one or a few but many phases of intergradation. Two marked leaf 

 types frequently occur: (a) leaf -lobes of the pinnately divided leaves narrow and long and entire 

 or sparingly laciniate-toothed or -lobed ; and (b) leaf -lobes short and broad and pinnately divided, 

 the sinuses rounded and the lobes mainly obtuse. Each leaf tjqie develops secondary types. Each 

 secondary type tends to be associated with two main corolla forms, one with long tube, one with 

 short tube. Brought into this complex also are plants with corollas of varying sizes, but varying 

 like length of corolla-tube independently of the leaves or other organs. Pubescence, glabrosity 

 and glandulosity are less easily describable but continuous variables, which like the other variables 

 are usually devoid of geographic correlations. 



As here described Gilia tenuiflora is an inclusive species, evaluated as an indivisible unit 

 within which variation is very great, continuous and ^vithout constant coordination or correlations 

 of variable factors. It is the kind of complex which is frequently spoken of by taxonomists as 

 "hopelessly confused," "deeply entangled" or "an aggregate of numerous hopelessly intergrading 



Fig. 381. GiUA TENUIFLORA Benth. 

 a, habit, X % ; b, fl., X IVz. Drawn 

 from Douglas type. Royal Botanic 

 Gardens, Kew (Herb. Benth.). 



