190 



POLEMONIACEAE 



prevailing form exhibits a close dense basal rosette and nearly naked stems. 2. The leafage, and 

 especially the lobation, of Ipomoposis inconspicua Sm. correspond to occasional collections of 

 Gilia miUefoliata. It is noteworthy that the upper leaves in Gilia millefoliata have often fewer 

 coarser lobes than the lower leaves and that this coarser type of leaf corresponds better to the 

 upper leaves in Ipomopsis inconspicua, the only ones shown in the illustration. 3. The loose in- 

 florescence of Ipomopsis inconspicua Sm. is well matched by such a form of Gilia millefoliata as 

 Samoa, Humboldt Bay, Tracy 1921. 4. The corolla of Ipomopsis inconspicua Sm. barely ex- 

 ceeds the calyx, just as in Gilia millefoliata F. & M., whereas in the tj'pe of Gilia inconspicua 

 Dougl. it is twice as long and in prevailing forms often 2 to 4 times as long. 5. The tube and 

 throat of the corolla in Ipomopsis inconspicua Sm. are broadly tubular, just as in Gilia miUefoliata 

 F. & M., whereas in Gilia inconspicua Dougl. the very slender tube abruptly expands into a broad 

 throat. 6. The calyx-teeth in Ipomopsis inconspicua Sm. are depicted as linear ; in the type and 



in the prevailing form of Gilia inconspicua Dougl. 

 the calyx-teeth are triangular-acute. Nothing 

 quite like the linear calj'x-teeth of Ipomopsis in- 

 conspicu.a Sm. is known either in Gilia inconspi- 

 cua or Gilia millefoliata, but in Gilia millefoliata 

 the teeth are commonly lanceolate. 7. The seed 

 of Ipomopsis inconspicua Sm. could not have 

 come from the untraveled interior of the Colum- 

 bia Eiver basin in that very early day; it is, as a 

 speculation, quite possible that it was obtained 

 on the California coast in 1792 by Archiljald 

 Menzies of the Vancouver Expedition. The seeds 

 collected by Menzies went first to England. 



It must be pointed out that in the region of 

 the type locality of Gilia inconspicua Dougl., col- 

 lections of this species, such as that made at 

 Lexington, llorrow Co., northeastern Oregon, 

 Leiherg 15, simulate rather closely in leafage and 

 somewhat in inflorescence the original illustration 

 of Ipomopsis inconspicua Sm., but the sum total 

 of the evidence, morphological and geographic, 

 leans towards the view that Ipomopsis inconspi- 

 cua Sm. is Gilia millefoliata F. & M. and this view 

 is here adopted. 



Since no type specimen exists for Ipomopsis 

 inconspicua Sm., the following lectotype is so 

 designated as answering most nearly the original 

 description and illustration: Samoa, Humboldt 

 Bay, Tracy 1921 (in Herbario Jepsoniano). The 



Fig. 386. Gilia tkicolor Benth. a, cauline 

 leaf, X V2 ; b, infl., X V-i ; c. fl., X 1%. Draivn 

 from Voualas type, Eoyal Botanic Gardens, 

 Kew (Herb. Benth.). 



following collections validate the range of Gilia 

 inconspicua Sweet : Smith Eiver Indian Eeserva- 

 tion, Del Norte Co., R. Van Deventer 309; Little 

 Eiver Beach, Tracy 4794; Samoa, Humboldt Bay, 

 Tracy 14,833; Bucksport, Humboldt Co., Tracy 

 2190; Fort Bragg, Mathews 55; Bodega Pt., Sonoma Co., Eastwood. 



Eefs. — Gilia inconspicua Sweet, Hort. Britt. ed. 1, 286 (1827), resting on Ipomopsis Incon- 

 spicua Sm., Exotic Bot. 1:25, t. 14 (1804), a garden plant cult, in England from seed believed to 

 be from "America." Ipomeria inconspicua Nutt., Gen. 1:125 (1818), resting on Ipomopsis incon- 

 spicua Sm. Cantua parvifiora Pursh, Fl. 730 (1814), resting directly on Ipomopsis inconspicua 

 Sm. G. parvifiora Spr., Sys. Veg. 1:626 (1825), resting on Cantua parviflora Pursh. G. mille- 

 foliata F. & M., Ind. Sem. Hort. Petrop. 5 :35 (1838), type loc. Bodega Bay, Sonoma Co.; Linnaea 

 13: Litt.-Bericht. 109 (1839); Erythea 2:164 (1894). G. multicaxdis var. millefoliata Jepson, 

 Man. 796 (1925), as to north coast plants. G. millefoliata var. maritima Brand; Engler, Pflzr. 

 4=^0:100 (1907), type loc. Humboldt Bay, Eastwood. 



14. G. tricolor Benth. Braos Eyes. (Fig. 386.) Stem slender, erect, with few 

 branches from or above the base, or with many branches from the base and ascend- 

 ing, 8 to 16 (or 21) inches high, more or less leafy on lower part, the thin basal 

 rosette of leaves usually disappearing early; herbage glabrate to somewhat glandu- 

 lar ; leaf-blades pinnate. 1 to 1% (or 31^) inches long, the petioles 14 to V2 as long, 

 the segments linear, mostly remote, entire or pinnately tootlied or laciniately cleft, 

 3 to 9 lines long; flowers 1 to 5 on short (i/^ to 2 lines) pedicels, borne in dense ter- 

 minal cymes, the cymes on slender peduncles; cah'x cylindrie to turbinate, some- 



