as Peucedanuni), C. tomentosa (Benth. PI. Hartw. 312 (18-W) 

 as Peucedaniim). C. marginata (Benth PI. Hartw. 312 (1849) 

 as Peucedaniim) a doubtful species. 



The second division of this section has those with caespit- 

 ose and mostly woody roots but thickened and somewhat 

 fleshy and branched below; plants nearly always growing 

 among- cliflFs in the clefts of rocks ; leaves mostly rather rigid, 

 ultimate segments sharp. The high alpine forms are C. Ore- 

 gana (C. & R. Rev. Umb. 64 (1888) as Peucedanum), C. Hal- 

 Hi (Watson Proc. Am. Acad. 11 141 (1876) as Peuceda- 

 num), C. Austinae (C. & R. Bot. Gaz. 13, 208 (1888) as 

 Peucedanum), is subalpine, C. Leibergii (C. & R. Umb. 224 

 (1900) as Lomatium), C. Sandbergii (C. & R. Bot. Gaz. 13 

 79 (1888) as Peucedanum). C. Congdoni (C. & R. Umb. 232 

 (1900) as Lomatium) and C. Piummerae (C. & R. 232 (1900) 

 as Lomatium) may come here. The low altitude forms of the 

 Lower Temperate and Tropical Life Zones are C. microcarpa 

 r Howell C. & R. Rev. Umb. 65 (1888) as Peucedanum), 

 C. millefolia f Watson Bot. Kin^ 129 (1871) as Peucedanum), 

 C. Torreyi (C. & R. Bot. Gaz. 14 276 (1889) as Peucedanuc), 

 C. Eastwoodae (C. & R. Umb. 247 (1900) as Cynomarathrum), 

 C. scabra fC. & R. Umb. 247 (1900) as Cynomarathrum). 



Cvnomarathrum Nuttall Coulter and Rose consider to be 

 a good genus but how they can separate it from the Ctisickii- 

 triternatum group they would have much difficulty in ex- 

 plaining. They say it differs from Lomatium 'in the more 

 Avingred dorsal and intermediate ribs" when L. alatum is fully 

 as much winged while anomalum has ribs three or four times 

 as much winged. They say it differs in its habit and yet they 

 would have a hard time telling in what the habit differs from 

 that of C. Cusickii or the triternatum group except that the 

 one grows among rocks and the other in crevices. It is true 

 the leaves are pinnate but so are those of Hallii, Oreganum 

 and Leibergii. In Nuttallii the seed face is not so flat as 

 in triternatum. The Eastwoodae section gradually passes into 

 the utriculatum section on the one hand and through Parryi 

 into the Nuttallii section, but they do not differ from either as 

 much as bicolor does from utriculatum, and still less from the 

 foeniculaceum or Mohavense section. Taking up Euryptera 

 they also make that a genus and yet it differs less from the 

 leiocarpum section than that does from the true Lomatia. It 

 is manifestly a close congener. To separate these groups into 



