258 GEOGRAPHICAL AND GEOLOGICAL DISTRIBUTION. 



deed, a very large proportion of the species appear to be restricted 

 to special zones of a given formation.* 



In their geographical relations the ancient Brachiopoda differ 

 essentially from those of the modern seas by reason of their broad 

 horizontal distribution. Cosmopolitanism, or something approach- 

 ing it, if not exactly the rule, was at least distinctive of a very 

 large proportion of the species, a circumstance undoubtedly due 

 to more equable conditions of environment, and the absence of in- 

 terposed barriers to migration. This broad dispersion is perhaps 

 nowhere better illustrated than in the case of the extinct fauna of 

 China. Thus, out of a total of thirteen Silurian and twenty-four 

 Devonian species described in Richthofen's work, no less than ten 

 of the former and sixteen of the latter are also found in Western 

 Europe; and, further, of the Devonian species, about eleven, or 

 nearly fifty per cent., are cosmopolitan. Again, of about twenty- 

 five Carboniferous species, some fifteen (or sixty per cent.) are 

 common to North America, and about an equal number are cosmo- 

 politan. 



MOLLTJSCA GENERALLY. 



The more salient features connected with the distribution of 

 the animals of this class have already been considered in our treat- 

 ment of Geographical Distribution and Brachiopoda, and do not 

 require restatement. It has been seen that the principal factors 

 involved in this distribution are temperature and the presence or 

 absence of continuous coast-lines along w r hich migration might be 

 effected; to these two categories may also be added the circum- 

 stance of light (or darkness), but to what extent this influence is 

 exerted has not as yet been determined. 



The attempted subdivision of the oceanic area into a number 

 of distinct regions (provinces), each one characterised by a more or 

 less peculiar assemblage of molluscan forms, while undoubtedly 

 indicating a certain amount of faunal individuality, is still far from 



* A more critical and impartial revision of the species will not unlikely 

 materially increase the number of such connecting forms, and forms of even 

 widely separated formations may be found to be identical. Thus, Davidson 

 ("British Jb'ossil Brachiopoda," 4t Palaeont. Soc. Rep.," 1884, p. 396-398) ad- 

 mits that some of the forms of the Mediterranean Tcrebratulina caput-serpentis 

 may only be varieties of the Cretaceous T. striata ; and, likewise, that the 

 recent Rhynchonella nicrricans and some Cretaceous and Jurassic forms are so 

 closely related to each other " that we are at a loss to define their differences." 



