men of which he enclosed in his letter. That the plant in question could not possibly belong to the genus Tagetes was shown at once by the form of the involucre subtending the flower-head, which according to Hernandez's drawing, was not at all cupshaped, or tubular like that of Tagetes but composed of several distinct linear bracts. A comparison of Hernandez's rude illustration (fig. 1) with herbarium specimens of Tagetes showed that the plant in question could not possibly be included in the same genus with them. His figure represents a composite with flowerheads not unlike those of a Coreopsis, but the accompanying leaves are artemisia-like as stated in the original description. The widely spread ray-flowers, few in number, are three-toothed at the apex. On one of the heads they have fallen off, indicating that they are not persistent like those of Tagetes. The disk flowers form an erect cylindrical bundle, while the entire head is subtended by an involucre not at all like that of the genus Tagetes, but composed of a few linear sepal-like bracts as stated above. Failing to find the plant in the genus Coreopsis, the writer carefully examined the plants belonging to allied genera. At, last, in the genus Cosmos, he came upon a species corresponding in all respects with Hernandez's figure. The long sought Xochipalli proved to be Cosmos sulphureus, a species which, though figured the latter part of the eighteenth century by two eminent botanists, had never been associated with the dye-plant described and figured by the great protomedico a century previously. To verify the discovery, a decoction of the flowers was made for the writer by Dr. L. A. Hawkins, Plant Physiologist, of the Department of Agriculture. Almost immediately the water became suffused with an orange tinge, and on the addition of a very small quantity of alkali it changed to a rich orange-red, the color, of the xochipalli described by Hernandez. Though never hitherto connected with the classic "flowercolor" used by the Aztecs in painting their codices, Cosmos sulphureus is not a rare plant. Specimens of it were lacking in the United States National Herbarium until 1886, when it was encountered by the veteran explorer Dr. Edward Palmer in the vicinity of Guadalajara, Jalisco; and five years later he collected it at Culiacan, Sinaloa, bringing back with him from this locality seeds from which plants were propagated at Washington. The account of its introduction into cultivation in the United States is told by Dr. J. N. Rose in Garden and Forest for December, 1895, where an excellent figure of it was published. It is now represented in the United States National Herbarium by specimens from many other parts of Mexico: from Durango, Sonora, Tepic, Colima, Oaxaca, Chiapas, Guerrero, and Morelos. It is interesting to note that this species, observed by Gemelli-Careri centuries ago while traveling between Acapulco and Cuernavaca, has been collected at both of these terminals, at Acapulco by Dr. Palmer, in 1894, and at Cuernavaca by Dr. J. N. Rose, in 1902. Seeds of this classic dye-plant of the Aztecs were recently obtained by the writer, and he now has a number of vigorous young plants of the true xochipalli growing in one of the greenhouses of the United States Department of Agriculture. DESCRIPTION Cosmos sulphureus is a tall, rank, pubescent annual composite, growing usually about four to seven feet high, with stems as thick as the thumb and bipinnatifid or tripinnatifid leaves, not unlike those of the common Artemisia vulgaris in form. The flower heads, borne on long slender peduncles, are subtended by a calyx-like involucre composed of two series of eight bracts each, the outer bracts linear and green, the inner broader and scarious. The flowers vary in color from bright orange to deep reddish orange. The heads are composed of eight broadly ovate ray-flowers, three-toothed at the apex, spreading at right angles to the axis and soon falling off, and fertile tubular disk flowers forming a compact erect cylindrical bundle. The exserted anthers are black with orange tips, and the style is branched, terminating in two slender tips. The fruit is a linear akene nearly an inch long, including the slender barbed beaks, and the pappus consists of two slightly hispid awns. REFERENCES Cosmos sulphureus Cavanilles, Icon. Pl. 1: 56, pl. 79. 1791. Rose, Gard. & For. 8: 484, fig. 66. 1895. Coreopsis artemisiaefolia Jacquin, Icon. Pl. Rar. 595; Col. Suppl. 155. 1791. Xochipalli folia ferens sinuosa, magna, et Artemisiae quadantenus similia. Hernandez, in Nard. Anton. Recchi, Rerum Medic. Nov. Hisp. Lib. 7, Cap. 36. 1651. In the accompanying illustration (fig. 2) the plant is shown, slightly reduced, accompanied by drawings of a disk flower (a) and an achene (b), the two latter enlarged about two diameters. BOTANY.-A new Polystichum from California.1 WILLIAM R. MAXON, National Museum. In reviewing recently the Polystichums of the Pacific coast region of the United States as represented in the National Herbarium and in the very ample collections of the Dudley Herbarium of Leland Stanford Junior University, the latter lent for study through the courtesy of Professor LeRoy Abrams of that institution, it has been found that the plant of the Santa Cruz peninsula and of two or three localities not far northward which has commonly been referred to Polystichum aculeatum (or formerly to its "var. angulare") represents a new species. This is described below, being named in memory of the late William R. Dudley, for many years professor of botany in Leland Stanford Junior University. Polystichum Dudleyi Maxon, sp. nov. Rhizome stout, decumbent, together with the numerous stipe-bases densely paleaceous, the inner basal scales mostly linear-lanceolate, attenuate, dark brown, dull, semiopaque, imbricate, those above spreading, larger and broader, narrowly oblong to ovate, long-acuminate, up to 2 cm. long, brown to fulvous, concolorous, translucent, membranous, distantly denticulate-ciliolate, extending well up the stipe but mostly deciduous above the basal third, underlaid by a dense series 1 Published with the permission of the Secretary of the Smithsonian Institution. of small, narrow, attenuate, paler scales, these freely denticulateciliate. Fronds several, erect-spreading, 40-120 cm. long; stipes stramineous above the dark arcuate base, usually more than half as long as the blade, sulcate; blades narrowly ovate to narrowly oblonglanceolate, very long-acuminate to long-attenuate, 25-75 cm. long, 8-25 cm. broad, bipinnate; rachises densely and subpersistently paleaceeous, the scales pale, mostly linear-attenuate, or acicular from a small, roundish-cordate, fimbriate base; pinnae numerous, the lower ones slightly apart, spreading, a little shorter than those above, these mostly contiguous, oblique, often strongly arcuate, linear to narrowly oblonglanceolate from a broader, invariably inequilateral base, attenuate from the middle, the apex rather abruptly acute to long-acuminate; pinnules numerous, oblique, straight or usually falcate, petiolulate, obliquely ovate or ovate-oblong from a distinctly cuneate, inequilateral base, the superior basal pinnule much the largest, this obliquely cleft or with several pairs of nearly free segments, the other pinnules auriculate, serrate to obliquely incised, the acuminate apex and the curved serratures rather abruptly short-awned; pinnules with a copious covering of pale, lax, sinuous, tortuous, filiform scales beneath, a few similar scales borne on the upper surface; leaf tissue bright light green, papyraceous, the course of the veins evident above; sori numerous, terminal or subterminal upon the first anterior branch of the oblique veins, or several pairs borne upon the simple pinnately arranged veinlets of the auricles and of the segments of the superior basal pinnules; indusia orbicular, membranous, copiously long-ciliate, the cilia flaccid, septate, frequently equaling in length the diameter of the body of the indusium. Type in the U. S. National Herbarium, no. 887829, collected "near bridge, Peter's Creek," Santa Cruz Mountains, San Mateo County, California, May 2, 1903, by W. R. Dudley. Three sheets with identical data are preserved in the Dudley Herbarium, which contains also numerous other collections by Professor Dudley of the same plant from the Santa Cruz peninsula, as follows: Near bridge below La Honda, San Mateo County, March 24, 1894; King's Mountain, San Mateo County, March 5, 1898; sandstone rocks, Upper Gazos Creek, San Mateo County, January 1, 1902; cliffs near Camp on Pescadero Creek, Santa Cruz Mountains, San Mateo County, June 19, 1905; rocky canyon of east fork of Waddell Creek, Santa Cruz County, September 23, 1901; south fork of Big Creek, Santa Cruz County, August, 1903; Los Gatos Canyon, Santa Clara County, April 21, 1895, July 22, 1895. Other specimens from the same or near-by regions are as follows: Foothills west of Los Gatos, Santa Clara County, Heller 7226; Hume Canyon, Saratoga, Santa Clara County, alt. 270 meters, Pendleton 1468; Santa Cruz Mountains, Anderson, Miss E. B. Norton, Mrs. A. E. Bush, Miss Bowles; San Gregorio, San Mateo County, May 31, 1870, Brannan & Kellogg; Ukiah, Mendocino County, May 11, 1869, Kellogg; above "Grimes," Santa Lucia Mountains, Monterey County, August, 1903, Dudley; without special locality, Kellogg & Harford 1183. |