1916 1 Grinnell: Analysis of Trinity Fauna 409 



vary as circumstances vary, just as with the recognition of genera, 

 families, etc. The term subfauna has been previously employed for 

 the more slightly characterized assemblages of animal inhabitants, 

 and this might be a useful term to adopt regularly in connection 

 with such minor faunal divisions as that of the Trinity region. 



The exact location of boundaries for the Trinity subfauna is a 

 matter of uncertainty. This uncertainty arises from two circum- 

 stances : first, that, as experience elsewhei-e would lead one to expect, 

 there are broad marginal belts of intermingling, rather than sharp 

 lines of demarcation ; and, second, lack as yet of exploration of much 

 of the surrounding territory. No field-work has been done by anyone 

 connected with the California Museum of Vertebrate Zoology on a 

 line directly to the coast west from the Salmon Mountains. To the 

 north, the Siskiyou Mountains, across the deep Klamath River canon, 

 are so far known only as regards their birds (see Anderson and Grin- 

 nell, 1903). The birds are indicative of close faunal similarity of 

 the Siskiyou and Trinity mountain masses. To the south, the Yolla 

 Bolly country presents some obvious peculiarities of its own, but 

 here, again, no adequate study has been accorded the animal life, 

 and conclusions are therefore best deferred. To the east, the Trinity 

 area seems satisfactorily bounded by the upper Sacramento Valley 

 and by Shasta Valley together with the interconnectant Sisson gap. 

 extending between mounts Shasta and Eddy. 



Summary. — The collections of specimens thus far available from 

 the Trinity region cannot be considered anywhere nearly complete ; 

 neither is there sufficient information at hand, published or otherwise, 

 concerning the Cascades. Even with the best-worked of the faunas 

 here concerned, that of the Sierra Nevada particularly in its Mount 

 Shasta portion, much remains in doubt. The statistics here set forth 

 must therefore be considered provisional. Nevertheless the general 

 conclusions are believed closely to approximate the truth. 



The Trinity region as regards its Boreal fauna is clearly far more 

 closely allied to the Sierra-Cascade fauna than to the Humid Coast 

 fauna. It is nearer to the Sierran fauna than to the Cascade fauna ; 

 indeed it might with propriety be included in the Sierra Nevadan 

 faunal area, ranking merely as a subfauna. 



The Trinity region contains some Sonoran "islands". The fauna 

 of these is most closely similar to that of the Sacramento Valley; 

 there are a few Great Basin or Modoc forms, and but scanty evidence 

 of humid coast intrusion. 



