522 GEOLOGY 



(1) Edaphic modification. Rocky, sandy, muddy, and cal- 

 careous bottoms had their appropriate life, as did also tracts of 

 shallow and deep water, and areas dominated by other special 

 conditions. The assemblages adapted to these special conditions 

 were not altogether unlike, for not a few forms, particularly free- 

 swimming species, were indifferent to these conditions. 



(2) Provincial modifications. Although the sea covered a large 

 part of the continent, affording facilities for the migration and 

 mingling of faunas, there was still evidence of some separation into 

 zoological provinces. This was probably due partly (1) to barriers 

 interposed by gentle warpings of the sea-bottom producing emer- 

 gent tracts and tracts of excessive depth, and partly (2) to barriers 

 constructed by the sea itself, in the form of shoals, bars, and spits. 

 Provinces may have been defined also by (3) ocean-currents with 

 their attendant differences in temperature, and they may have been 

 due (4) to variations in the salinity of the waters. Provinces due 

 to surface warpings seem to have been most marked in the Appa- 

 lachian tract. 1 



(3) Cosmopolitan development. Notwithstanding the local and 

 provincial modifications just noted, the progress of the Ordovician 

 life on the American continent seems to have been, on the whole, 

 in the direction of cosmopolitanism. This was due, primarily, to 

 the wide development of the epicontinental seas, which gave a 

 broad field for the evolution of marine life, and permitted free 

 migration. A cosmopolitan tendency is particularly marked in 

 the great interior of the continent. These statements apply 

 chiefly to the shallow-water faunas. The deep-sea beds of the 

 period are inaccessible. 



The Ordovician system contains an exceptionally large niiinluM- 

 of fossils of free-floating graptolites (Fig. 394) . 2 Their remains arc 

 mingled with the fossils of the shallow-water life, showing that 

 pelagic life swam freely over the epicontinental seas. The Ordo- 

 vician graptolites are nearly identical in Europe, North America, 



1 Paleozoic Seas and Barriers; E. O. Ulrich MIX! Chnrlcs Srhuchcrf . Kept. 

 N. Y. State Paleontologist, 1901, pp. 633-658. 



2 It is not universally agreed that all graptolites were floating forms at all 

 stages, but there seems to be little doubt that they usually urn- in their 

 young stages at least. 



