222 University of California Publications in Botany [^'ol. 10 



1893; alonjr creek near Yreka. Butler 912, 389; Yreka Creek, Butler 

 1610 (form with leaves and lei;nme.s densely villous) ; Edgewood, K. 

 Brandegee, July, 1887 ; bank of Mad River, CJiesnut and Drew, July, 

 1888; Jarnijran's, Humboldt County, Chesnut and Drew, Julv 11, 

 1888; Mad River, Pr^Vr, July ], 1890; Harris to Fort Seward Ranch 

 Ridge, Humboldt County, Jepson 1882 ; Kneeland Prairie, Tracy 3845 ; 

 valley of Van Duzen River, Tracy 2756 ; near Hydesville, Tracy 1240 ; 

 Chico-Hamilton road, six miles from Chico, Heller 11342 ; Tehachapi, 

 Kern Co., Greene, June 21, 1889 (out of range but stem, calyx teeth, 

 and legumes tj^pical L. denticulatus). 



References. — Lotus denticulatus Greene, Pitt., vol. 2, 1890, p. 

 139. Hosackia denticulata Drew, Bull. Torr. Club, vol. 16, 1889, p. 

 151, type loc. along Mad River, near Jarnigan's, Humboldt County, 

 Chesnut and Drew. Anisolotus denticulatus Heller, ]\Iuhl., vol. 7, 

 1912, p. 139. 



17. L. subpinnatus Lag. (pi. 74, figs. 15-20). Closely resembling 

 L. denticulatus Greene in many of its technical characters and often 

 distinguished from it with dit^iculty in herbarium specimens; low, 

 diffusely branched, pilose to nearly glabrous; leaflets 3 to 5, mostly 4; 

 flowers yellow, tinged with reddish purjDle ; calyx teeth usually shorter 

 than the tube; ovary oblong, less hairy than for L. denticulatus; 

 legumes 3 to 7-seeded ; seeds 2 mm. long by slightly less than 1.5 mm, 

 wide, shape as in L. denticulatus. 



More or less common in the Sonoran throughout the state. Often 

 found growing side by side with L. humistratus Greene. Distribution 

 Map 6. March-October. Upon comparison of certain Calif ornian 

 specimens of the U.C. Herbarium {TIaU 3841, Brewer 360, 1166, see 

 localities) with plants at Kew Herbarium of Lotus suhpinnatus Lag. 

 collected in Chile, Dr. Jepson found that, in general, the Chilian 

 plants were more pubescent, the leaflets were narrower and were four 

 in number, not five as is often the case with the Californian plants. 

 After a careful study of a large number of plants from different parts 

 of the state and a consideration of the diagnosis of the Chilian type 

 by Lagasca and of the description given by Hooker and Arnott for 

 the species as represented by plants collected at Conception, Chile, 

 I have come to the conclusion that the Californian representatives 

 should be retained within the species. They show as great variation 

 among themselves as with the South American plants. The South 

 American plants are also variable, as may be seen from a comparison 

 of Hooker and Arnott 's description Avith that of T^agasca. The extent 

 of variation is no more than one might expect to find in a species with 

 such a wide range of distribution. 



Localities. — Near Trinidad. Humboldt County, Tracy, 2967 ; 

 Yager Creek near Hydesville, Tracy 2521 ; T'kiah, Purdji April 13, 

 1898 ; Fort Ross, Sonoma County, Heller 6599 ; Bodega Point, East- 

 wood, June 9, 1899; Bodega Bay, Chandler 719; Vanden and Little 

 Oak, Jepson in 1885 ; near Napa, Heller and Brown 5358 ; Ross Vallev, 

 Marin County, Chandler 292; Fairfax, Ottley 360; Holiday Hill, San 

 Francisco, Rattan in 1887; Lands End, San Francisco, Davy, June 

 12, 1896; vicinity of Berkeley, Walker, 56; North Berkelev Hills, 

 Chandler, March 17, 1900; Berkeley, Chandler, 243; U. C. Campus, 



