392 J. BARRELL INFLUENCE OF CLIMATES ON VERTEBRATES 



Both marsipobranchs and fishes show a ready adaptation to the differ- 

 ences of salinity between fresh and salt waters. In many species the 

 individuals pass from salt to fresh waters for breeding purposes, certain 

 fishes, such as the eels, possessing, however, the opposite habit. The 

 greater safety of fresh waters as an environment for the young is a suffi- 

 cient explanation of this common habit among fishes; yet the ready 

 adaptation to different salinities in contrast to the narrow limits common 

 in other phyla of animals is suggestive that even in early times chordates 

 were not limited in their range by the low or high salinity of the waters, 

 but by the enemies of the one or the other environment. 



The habits of the lampreys are most suggestive in this connection. 

 These lowest of craniate vertebrates live both in fresh and salt water and 

 ascend the streams to spawn. They are effective swimmers, capable of 

 stemming the currents of swift rivers and using the same power to over- 

 take the fishes which they make their prey. The only known fossil rep- 

 resentative, Paleospondylus gunni from the Devonian, was found, further- 

 more, in a fresh-water formation — the Old Eed Sandstone of Caithness. 

 Although an apparent and perhaps sufficient cause of the spawning habits 

 of the lampreys is to be found, as in true fishes, in the greater safety of 

 the larva?, yet in such an archaic form it is likely to be an equally archaic 

 habit. The more archaic the habit, the more likely it is to signify that 

 the fresh waters were the environments within which the lampreys came 

 into existence. In the same manner, it is seen that the laying of the eggs 

 of amphibians in fresh water is an embryological necessity reminiscent of 

 their original home before their race grew to possess lungs and legs. 



The diversification of true fishes, when they assume a dominant role 

 in the Devonian, points to a long previous evolution, yet the marine record 

 of their previous existence is almost negligible. Compare the abundant 

 and well preserved fossils of brachiopods, trilobites, and cephalopods in 

 the Cambrian and Ordovician rocks with the rare and fragmentary evi- 

 dences of fishes. Yet the teeth, spines, and plates which ostracoderms 

 and acanthodian sharks possessed in the Ordovician should have per- 

 mitted ready fossilization if representatives of these subclasses were then 

 broadly present in the sea. 



The lowest well known occurrence of fishes in the stratigraphic column 

 is that made known by Walcott in 1891. Near Canyon City, Colorado,, 

 the base of the Middle Ordovician (Lower Trenton) rests on the Pre- 

 cambrian. The basal formation is the Harding sandstone, 86 feet thick. 

 It is overlain by from 2 to 4 feet of shale, followed by about 300 feet of 

 limestone holding an abundant Trenton marine fauna. The Harding 

 sandstone is without doubt an offshore formation, since it contains species 



