NO. 2 OLGA hartman: submarine canyons 7 



confluent with the Santa Monica basin at their lower ends. San Pedro 

 sea valley terminates in the San Pedro basin. Newport, La Jolla and 

 Coronado canyons merge with parts of the San Diego trough. The 

 offshore canyons, from Santa Cruz at the north, to San Clemente 

 rift valley at the south have the following terminations: Santa Cruz 

 canyon merges with a basin of the same name, as do also Catalina and 

 San Clemente canyons. Tanner cannon merges with the East Cortes 

 basin. The lower ends of these canyons consistently share the biological 

 features of the basins which they join. Those terminating in Santa 

 Monica and San Pedro basins are nearly or altogether dead in their 

 basin subsill depths. Those farther south have species to be found also 

 in the San Diego trough. 



The outermost canyons differ in most respects not only from ad- 

 jacent ones, but also from longshore canyons. Specific limitations may 

 be imposed not only by geography, location of canyon, but also by the 

 character of its surface sediments (vs'hether mud, sand, rock), re- 

 moteness from land areas and available food supplies or from sources 

 of larval replenishment. Each of the canyons illustrates the fact that 

 there are abrupt differences in faunal components inter- and intra- 

 canyon-wise for which explanations are wanting. A far greater samp- 

 ling program in most canyons might result in a resolution of these 

 differences, if species now apparently lacking were found more widely 

 distributed. On the other hand, it might show even greater differences 

 than the present Analyses, if a much larger number of species were 

 added. 



Some conclusions can be stated respecting the faunas of the 

 canyons : 



( 1 ) Each canyon is found to support a richly diversified fauna, 

 high in specific entities, with as many as 262 species in a longshore 

 (Newport) canyon. 



(2) The largest numbers of species in a canyon occur in shal- 

 lowest, or shelf depths, and they are members of the shelf or slope 

 fauna. There is a gradual decline in numbers of species (though not 

 necessarily specimens) with depth^ but there are deviations from this 

 principle, perhaps partly due to factors other than depth. 



(3) Most species occur as single or few specimens in a sample, 

 shown by the frequent recurrence of the number 1 in the ANALYSES 

 (see below) ; these unit numbers apply to most species in most canyon 

 depths. 



