ENVIRONMENTAL CONDITIONS 393 



of Lingula, Orthoceras, and Beyrichia. At several levels water-worn 

 scales and plates of fishes are found abundantly. Many of these plates 

 belong clearly to an ostracoderm of the family Asterolepididae. One por- 

 tion of the head carapace was found. Another form,, based on a calcified 

 chordal sheath, Walcott provisionally assigned to the Chimaeroidea ; and 

 another form, based entirely on separated scales, he placed within the 

 crossopterygian family of Holoptychididas. These latter two forms, how- 

 ever, have been assigned with a question mark by Eastman to the ostraco- 

 derms. With the coming in of the limestones all traces of the fishes 

 disappear. 4 Similar fragmentary remains have since been found in some 

 other localities in the Bocky Mountains at the same geological horizon. 

 The indications are thus clear that the fishes at that time were already in 

 existence and possessed structures well adapted to fossilization, but were 

 not a part of the fauna of the open sea. Where could they have had their 

 previous existence and where was the central home in which they con- 

 tinued to exist until they again appear in the geologic record ? In answer, 

 it is seen that the thickness of the Harding sandstone suggests that it 

 consists of river-derived muds and sands. The worn fragments of fishes 

 were brought apparently to the margin of the sea by rivers and were 

 buried in this manner in a marginal marine deposit. 



One otherpositive trace of fishes is known in the Ordovician. Traquair, 

 writing in 1900, states: 



"Rohon has described from the Lower Silurian (Ordovician) of the neigh- 

 borhood of St. Petersburg small teeth {Palxeodus and Archodus) associated 

 with conodonts, and which seem to be real fish teeth, but not of selachians, as 

 is shown by the presence of a pulp cavity surrounded by non-vascular dentine. 

 It. is impossible to say anything more of their affinities." 5 



The next appearance of fishes is by means of the scattered spines known 

 as ichthyodorulites and referred to the genus Onchus. These spines be- 

 long to primitive sharks and occur in the Middle and Upper Silurian. 

 Claypole found them in Pennsylvania; but they can hardly be regarded 

 as declaredly marine, since, though possessing some marine associations, 

 ihey are found in red shales which a little farther east have the fades of 

 subaerial delta deposits, but are there entirely unfossiliferous. Their age 

 is probably Salina. 



The first notable assemblage of fishes is found in the stratum of fossil 

 remains known as the Ludlow bone bed. This occurs at the top of the 



* C. D. Walcott : Preliminary notes on the discovery of a vertebrate fauna in Ordo- 

 vician strata. Bull. Geol. Soc. Am., vol. 3, 1892, pp. 153-172. 



5 R. H. Traquair : The bearings of fossil ichthyology on the problem of evolution. 

 Geol. Mag., Decade 4, vol. vii, 1900, pp. 464, 465. 



