POPPY FAMILY 569 



Francisco, Gardner 526-531. These latter pass gradually into the form having erect branching 

 leafy stems 1 to l^/, feet high ivith large golden yellow' flowers (Lake Merced, San Francisco, 

 Gardner 17e), which is indistinguishable from and culminates in the var. crocea, the common 

 form further inland. In such a series there is no break and neither is any one character 

 constantly associated with any other character. In some cases the stems are somewhat slender 

 and ascending and but few in number (Cow Creek, Shasta Co., Baker ^ Nutting in 189i; Del 

 Monte, Monterey Co., Elnier 3551), while in other cases they are stouter and stiffly erect, and 

 numerous (near Walker Mt., Mendocino Co., Davi/ 1028 = var. crocea; near Copperopolis, 

 Calaveras Co., Davy 1396^ var. crocea); the root, while often thickened above may lose its 

 multicipital character (Cow Creek, Shasta Co., Baker # Nutting in 1894; Smith Sta.,"Yosemite 

 Park, Evans 1; Poso Creek, Kern Co., Hall # Babcock 5008) ; the basal and often the cauline 

 leaves may have long slender petioles and narrow elongated segments and the flowers long 

 peduncles (Eeche Canon, San Bernardino, Parish 5607). 



2. Glaucescence and pubescence. — Certain of the coast forms are not at all glaucescent 

 (Pt. Isabel, Contra Costa Co., Davy, Apr. 1897; Alton, Humboldt Co., Tracy 3765). In other 

 forms this character is sparingly evident (clay hills nortli of Ocean View, K. Brandegee loe; 

 Twin Peaks, San Francisco, K. Brandegee 14e; creek north of Mill Creek, Santa Lucia Mts., 

 Jepson 2610) ; while the extreme is reached in the insular and mainland coast form, var. mari- 

 tima Jepson (Surf, Santa Barbara Co., K. Brandegee 6e, 7e, 8e; Pt. Pinos, Monterey Co., 

 Heller 6539; Seaside near Monterey, K. Brandegee 2e, 3e). Pubescence, if present at all, is 

 itself variable in character, being rather fine in certain plants which answer to E. foeniculacea 

 Greene (Salinas Eiver near Castro^ille, K. Brandegee 12e, 13e) and of a scabrous nature in var. 

 maritima Jepson (Surf, Santa Barbara Co., K. Brandegee, 8e, 9e). 



3. Calyptra. — In typical E. californica (as figured by Chamisso) the calyptra is ovoid- 

 conical, 7 lines long, with a stout blunt beak about 2 lines long. From this form, however, 

 there is every intergrade to long slender beaks or short thick ones or even to nearly beakless 

 calyptras. Even in one collection or even on one individual there are frequently found beaked 

 and beakless calyptras: clay hills north of Ocean View, K. Brandegee 15e; Mt. Tamalpais, 

 K. Brandegee 37e; Gold Eun, Placer Co., K. Brandegee 38e; Tehachapi, K. Brandegee 39e. 



•1. Corolla. — In typical E. californica (as figured by Chamisso) the obovoid petals are 14 

 lines long. Specimens from the type locality vary considerably, the corolla in some plants being 

 scarcely more than % inch long (Twin Peaks, San Francisco, K. Brandegee 14e), whUe in 

 others it is 1% to 1% inches long (San Francisco, Bioletti in 1891). The seasonal variation 

 in size and color of the corolla in the var. crocea has been referred to above. 



There is likewise a wide range of variation in color in plants from the type region of 

 E. californica Cliani. Plants with flowers of a very deep orange color are frequently found 

 (clay hills north of Ocean View, K. Brandegee 15e ; Lake Merced, San Francisco, K. Brandegee 

 17e), while plants of the following series by Gardner are alike not merely in technical char- 

 acters but in habit and size, as alike indeed as if taken from the same root, and yet show 

 tlie following color variation: vieinitj' of Lake Merced, San Francisco, Gardner 525, "petals 

 clear light yellow, no evidence of orange"; 526, "clear light yellow except for orange spot 

 in claw"; 527, "clear light yellow witli definite deep orange blotch on lower third of each 

 petal"; 528, "petals, lower half deep orange, upper half light yellow"; 529, "petals deep 

 orange except for narrow fringe of light yellow"; and 531, "petals deep orange throughout." 

 In occasional instances the petals are yellow tinged with red: Presidio (serpentine outcrop), 

 San Francisco, K. Brandegee, May -June, 1908; Berkeley, E. A. Walker 201, Jepson June 21, 

 1905. In form and in number the petals are fairly constant, but there is occasionally a strong 

 tendency to variation : Simpson 's Ranch, Sweetwater Creek, Eldorado Co., K. Brandegee, 

 June 1, 1908, "locality about 15 X 20 m., plants in close proximity" (petals truncate or 

 rounded, entire or erose or with an abrupt tooth or scallop at tip, in some eases numbering 6, 

 the inner 3 narrower than the outer); Sonoma Valley, Jepson 4188 (corolla often 6-merous). 



5. Pistil. — The characters of the stigmas are scarcely used in either Greene 's or Fedde 's 

 key as forming any part of their basis for distinguishmg species, although E. xylorrhisa Greene, 

 E. scariosa Greene, E. gigas Fedde and E. ohartacea Fedde are described as having respectively 

 2, 8 or 10, many (12?), and 3 (or 6?) stigmas. In most of the other segregates the stigmas, 

 if mentioned at all, are said to be " 4 and unequal. ' ' The examination of herbarium material 

 shows that tliere is a tendency toward variation from 4 stigmas but that this variation is 

 not correlated with any other character : Presidio, San Francisco, K. Brandegee 20e, ' ' styles 

 2-4 on same plant"; Berkeley (about vacant lots, Oxford and Center sts.), K. Brandegee, 

 Nov., 1917; Grouse Creek, Humblodt Co., Cbesnut 4' Drew, Aug. 1, 1888; New York Ravine, 

 Eldorado Co., K. Brandegee, June, 1908; Parkfleld, se. Monterey Co., K. Brandegee, May 12, 

 1916. 



To sum up we find frequent evidence of fluctuating variability in a single collection or 

 even on a single plant, examples of it being apparently as mdespread as the species. It is 

 conceivable and indeed probable that tlie descriptions of the segregates of Greene and of 

 Fedde noted above, in each case resting upon a single specimen, may be but examples of 

 inconstant variation. 



