0. C. Marsh— Vertebrate Life in America. 841 
and Amia, the Dog-fish of our western lakes, which by their 
Structure are seen to be remnants of a very early type, are 
well represented by species so closely allied to them that only 
an anatomist could separate the ancient from the modern. In 
the succeeding beds, these fishes are still abundant, and with 
them are Siluroids nearly related to the modern Cat-fish 
(Pimelodus). Many small fishes, allied apparently to the 
modern herring (Clupea), left their remains in great numbers 
in the same deposits, and, with them has been recently found 
a land-locked Ray (Heliobatis). 
The almost total absence of remains of fishes from the Mio- 
cene lake-basins of the West is a remarkable fact, and perhaps 
may best be explained by the theory that these inland waters, 
like many of the smaller lakes in the same region to-day, were 
so impregnated with mineral matters as to render the existence. 
of vertebrate life in them impossible. No one who has tasted 
such waters, or has attempted to ford one of the modern alkaline 
lakes which are often met with on the present surface of the 
same deposits, will doubt the efficiency of this cause, or the 
easy entombment of the higher vertebrates that ventured within 
their borders. In the Pliocene lake-basins of the same region, 
remains of fishes were not uncommon, and in some of them are 
pliocene fishes are essentially those of to-day. _ : : 
In this brief synopsis of the past ichthyic life of this Conti- 
known in this country, are from the lower Devonian ; but these 
old fishes show so great a diversity of form and_ structure, as 
It is safe to infer, from the knowledge which we now possess 
of the simpler forms of life, that even more of the early fishes 
were cartilaginous, or so destitute of hard parts as to leave no 
enduring traces of their existence. Without positive knowledge 
of such forms, and considering the great diversity of those we 
have, it would seem a hopeless task at present to attempt 
to trace successfully the genealogy of this class. One line, 
however, appears to be direct, from our modern Gar-pike, 
through the lower Eocene Lepidosteus to the Lepidotus of the 
_ Cretaceous, and perhaps on through the Triassic Ischypterus 
