186 



BULLETIN 50, UNITED STATES NATIONAL MUSEUM. 



Southern Sonora (north to Gua3'mas ou the coast) and southwestern 

 Chihuahua (Batopilas, etc.). 



(hrpodacus frontalis rhodocolpus (not Carpodacus rhodocolpus Cabauis) Belding, 

 Proc. U. S. Nat. Mus., vi, 1883, 343 (Guaymas, Sonora). 



Carpodacus frontalis (not Fring ill a frontalis Hay) SALvixand Godman, Biol. Centr.- 

 Am., Aves, i, 1886, 421, part (Guaymas). 



CARPODACUS MEXICANUS RUBERRIMUS Ridgway. 

 SAINT LUCAS HOUSE FINCH. 



Siinilar to O. w. sono?'lensis, but wing and tail shorter and ])ill 

 larger. 



Adult wa/f'.— Length (skins), 128.27-148.08 (135.13); wing, 72.14- 

 74.93 (73.91); tail, 56.39-62.99 (58.93); exposed culnien, 10.16-11.18 

 (10.67); depth of bill at base, 9.65-9.91 (9.78); tarsus, 16.76-17.53 

 (17.27); middle toe, 12.70-13.46 (12.95).^ 



Adult fe?nale.—ljength (skins), 119.38-129.54 (124.46); wing, 68.58- 

 76.20 (71.63); tail, 50.80-57.40 (54.61); exposed culmen, 10.16-10.92 

 (10.41); depth of bill at base (one specimen), 12.45; tarsus, 17.02- 

 17.78 (17.53); middle toe, 12.70-13.97 (13.21)."^ 



This is decidedly the smallest of the several geographic forms of 

 Carpodacus mexicanuH^ only that found immediately aci'oss the Gulf 

 of California (6^. rti. ^so;^oW(e?i«^s) closely approaching it in measurements. 

 From the latter this peninsular form differs, so far as I can see, onlj^ 

 in its shorter wings and tail and slightly larger bill, the coloration 

 being quite identical. In this latter respect both these allied forms 

 are scarcely, if at all, different from C. m. rJiodocoJpus, of the south- 

 western l)order of the Mexican plateau, but their decidedly smaller 

 size will serve to readil}' distinguish them. 



Carpodacus frontalis (not Fringilla frontalis Say) Baird, Proc. Ac. Nat. Sci. Phila., 

 1859, 301, 304 (Cape St. Lucas; crit.) .—Cooper, Cm. Cal., 1870, 156, part (Cape 

 St. Lucas). — Salvin and Godman, Biol. Centr.-Am., Aves, i, 1886, 421, part 

 (Lower California) . 



although the series is too small to show whether this is actually the case or not. The 

 specimens examined average as follows: 



'Eight specimens. 



' Four specimens. 



