LOWER CRETACEOUS FORMATIONS AND FAUNAS 609 
European Cretaceous, but there are some variations, such as the 
occurrence of Lytoceras and Phylloceras with boreal forms in 
the Shasta and their absence from the southern Comanche where 
they belong. 
As to the causes of the sharp contrasts between the Creta- 
ceous faunas of the Texan region and the Pacific Coast, reasons 
based on the distribution of the fossil plants and on the geo- 
graphic relations of the deposits have already been given for 
denying any considerable influence to climate, that is, difference 
in temperature. Dr. White has argued for the existence of a 
long, narrow, continental land barrier between the two areas of 
deposition, and that is doubtless the only single cause that seems 
at all adequate to explain the facts. But how far south the barrier 
extended and whether its position was constant we do not know. 
It is known that the Comanche fauna at one time extended 
almost to the present Pacific Coast in northern Sonora, and there 
is evidence that the Pacific fauna may have extended as far east 
as Catorce in San Luis Potosi, but whether the two seas were 
really connected or what their exact relations were we can only 
hope to learn when the details of Mexican geology are fully 
studied. Besides the existence of a barrier, the depth and clear- 
ness of the sea and the character of the bottom also probably 
had considerable influence. These conditions would at least 
explain the abundance of echinoids in one case and their 
absence in the other. The Shasta beds give every evidence of 
rapid deposition near shore in shallow waters, with usually a 
muddy bottom, while a large part of the Comanche series was 
laid down in deeper waters, or at least little influenced by clastic 
deposits derived directly from the land. One other factor has 
suggested itself, though it is probably not of much weight, and 
that is that the gregarious mollusks, like Aucella and the 
Ostreidz, actually monopolized the sea bottom where either 
became well established, so that there was not room for both in 
the same area. 
The study of these two incomplete faunas should emphasize 
the danger in depending on the statistical method in correlating 
