ee ee ee ee 
COULTER AND ROSE——NORTH AMERICAN UMBELLIFERAE. 31 
Carpel turgid, becoming depressed on the back, with neither ribs nor 
oil tubes; the whole dorsal region inflated, the seed cavity being on 
the commissural side of the carpel section. Seed flattened dorsally; 
the face and back, plane or convex. 
Slender branching annuals, with stellate pubescence, opposite simple 
(lobed) leaves, scarious lacerate stipules, and simple few-flowered 
umbels of white flowers on axillary peduncles. 
Type species, 2B. palmata Ruiz & Pavon, Fl. Peruv. 3:28. pl. 251. 
1802. 
A genus of about 20 species, chiefly South American. 
Our generic characters are drawn chiefly from our own species, and do not entirely 
apply to all the South American forms, 
1. Bowlesia septentrionalis VU. & R.. sp. nov. Fic. 3. 
Weak, 5 cm. to 3 or even 6 
dm. long, dichotomously branch- 
ing; leaves thin cordate to reni- 
form, 1.5 to 8 cm. broad, 3 to 
5 lobed (lobes entire or toothed), 
on long, slender petioles; um- wa 
bels 1 to 4 flowered, on short _ ‘ 
peduncles; fruit about 2 mm. Fig. 3.—Bowlesia septentrionalis: a, b, x8. 
long, sessile or nearly so. 
Type locality, near Tucson, Arizona; collected by Myrtle Zuck, May 
16, 1896; type in U. 5. Nat. Herb. 
From Texas to southern California and north to the Sacramento 
Valley. 
Specimens examined: 
Texas: Mex. Bound. Surv. 410; near Austin, Hall, May 18, 1872; near Galveston, 
Joor, April, 1877; San Antonio, Havard, March, 1884; Round Rock, Bodin, 
January, 1890; Austin, Bodin, in 1891; Corpus Christi, Heller 1493, March 
23-30, 1894. 
Arizona: Verde Mesa, Smart 128, in 1867; banks of the Rillita, Pringle, April 
10, 1881; Tucson, Parish, April, 1884; same station, Toumey 193, April 12, 
1892; type specimens as cited under type locality. 
JALIFORNIA: Monterey, Mer, Bound. Surv., in 1850; near Fort Tejon, Xantus de 
Vesey, in 1857-58; Bolander 4633, in 1866, Humboldt, Kellogg & Harford 297, 
in 1868-69; San Diego, Jones 3065, March 14, 1882; Santa Monica, Hasse, 
April, 1891; Caliente, Kern County, Brandegee, April 4; Drytown, Ama- 
dor County, Hansen 1561, April 8, 1896; near San Jacinto, Leiberg 3221, 
April 1, 1898. 
Urban, in Flora Brasiliensis, states that the above forms are not B. lobata Ruiz & 
Pavon, as they have long been called, but refers them to B. tenera Spreng., which he 
makes a variety of B. incana Ruiz & Pavon. In many ways our plants seem nearer 
to B. tenera Spreng. than to any other described species, but that is a Brazilian form, 
They were taken by Nuttall to represent a good species, and are so labeled in the 
Philadelphia Academy, but his name is not now tenable. 
