INTRODUCTION AND SUMMARY 389 



In contrast to the comparatively continuous and widely developed 

 mantle of marine sediments of the early Paleozoic with its wealth of 

 fossils, the record of the land waters is so meager that practically nothing- 

 is known of its life. With the opening of the Devonian, however, the 

 formation of intermontane basins, especially in northwestern Europe, 

 gave rise to mixed river, swamp, and lake deposits shut off from the sea. 

 There were thus created physical conditions suitable for the preservation 

 of fossils of the fresh waters. Here, in the Lower Old Eed Sandstone, 

 numerous species of ostracoderms and acanthodian sharks are preserved, 

 The record shows that fishes were at this time well represented in the 

 fresh waters, while hardly as yet able to survive in the sea. 



At the opening of the Devonian the acanthodian sharks were the domi- 

 nant fishes in the continental waters. Dipnoans and crossopterygians 

 were not in evidence in the regions where the fossils have been preserved. 

 In the Upper Devonian the sharks were gone from the fresh waters, but 

 the crossopterygians had risen in that environment to a dominant place. 

 Their ability to utilize air adapted them better as river fishes to warm 

 climates marked by alternation of wet and dry seasons. It was a semi- 

 arid rather than an arid or humid climate which characterized the regions 

 of the Old Eed Sandstone during late Silurian and most of Devonian 

 time. The climate of much of. the torrid zone at the present time may 

 give the best illustration of the conditions of temperature and rainfall 

 which in the Silurian and Devonian, but especially in the later Devonian, 

 characterized the northern hemisphere. The few remaining dipnoans 

 and crossopterygians still live as fresh-water fishes under similar climatic 

 conditions. 



The amphibians are represented in the later Paleozoic by certain skele- 

 tons preserved in the accumulations of coal swamps, but more abundantly 

 by footprints in formations having the character of semi-arid floodplain 

 deposits. As the record is traced backward, the skeletons disappear, and 

 the oldest abundant traces are footprints left in shales and sandstones, 

 chiefly red in color, of Lower Carboniferous age. These are the deposits 

 of rivers which were in the main subject to seasonal shrinkage in mark- 

 edly' semi-arid climates. The footprints lead us back to the habitats of 

 river fishes, the ancestors from whom they sprang. There are here con- 



Paleozoic fishes. The papers of Grabau and the writer, although similar in conclusions, 

 are mostly supplementary in nature to each other, as is apt to be the case with inde- 

 pendent work. Grabau enlarges especially on the American fish faunas, and regards 

 even the Middle and Upper Devonian faunas of the central States as not yet completely 

 marine, although the fossils he notes are found in marine associations. This is a prob- 

 lem somewhat aside from the present paper. Consequently, although the present writer 

 has accepted what seemed to him more probable, a truly marine existence for fishes 

 from the beginning of the Middle Devonian, the difference of opinion on this topic does 

 not involve any of the principles under discussion in this paper. 



