1R»6.] Life Before Fomh. 283 



If we may try and rebuild this vanished beast of prey from 

 conjecture, aided by collateral evidence, we should consider it 

 an elongated, flexible form, developed from some swimming 

 worm-like ancestor, perhaps like the Ascidian embryo, stiff- 

 ened internally by a cord of firm flesh extending lengthwise 

 through the body, and moving not by cilia, but by the aid of 

 fleshy side flaps, the progenitors of the fin. We conjecture it 

 to have been, in short, the early stage of the fish, a creature 

 perhaps of considerable size and strength, due to the abund- 

 ance of easily obtained food, but as destitute of hard parts and 

 as little likely to be fossilized as Amphioxus. 



We may offer this conjecture with some safety, for it is not 

 long before we come upon actual traces of fish, and of a degree 

 of development which indicates a long preceding stage of evo- 

 lution. In fact, the fish in time appears to have been forced 

 to put on armor, as its prey had earlier done. Internicine war 

 began in the fish tribe itself. A wide specific variation arose, 

 with great differences in size and strength, the stronger at- 

 tacked the weaker species, and eventually two distinct types of 

 fish appeared, the Elasmobranch and the Ganoid ; the former, 

 represented to us by the modern sharks, being much the most 

 powerful and voracious, and holding the empire of the open 

 seas, while the latter dwelt in' shallower waters. The Ganoids, 

 preying on the bottom forms, become themselves the prey of 

 their strong and active kindred, and, as a result, the evolution- 

 ary process just described was resumed. The weaker fish put on 

 armor, in many cases heavy and cumbrous, a dense bony cov- 

 ering which must have greatly reduced their nimbleness, but 

 which safety imperatively demanded. It is these armored 

 forms that first appear to us as vertebrate fossils; the first fish, 

 as the first mollusk or crinoid known to us, being the resultant 

 of a very long course of development. As regards the Elasmo- 

 branchs, they, too, became in a rc 

 sufficiently to indicate any very t 

 selves. 



There is little more which we can say in this connection. 

 The story of the evolution of life bears an analogy worth men- 

 tioning to that of the development of arms of offense and de- 



