THE DEVONIAN FISHES OF MISSOURI 71 



side of the river. This reference seems to have been made from 

 a field examination of the fossils.* 



In the Providence region a green sandstone two to eight 

 inches in thickness overlies the Devonian unconformably and 

 underlies the Chouteau with apparent conformity. It has 

 been called the fish bed on account of the large number of frag- 

 ments of teeth of Ptyctodus calceolus that occur in it. Near 

 Fulton this sandstone gives place to a conglomerate that also 

 bears fish teeth in abundance, and in Warren and Montgomery 

 counties a red sandstone, usually one to five feet in thickness 

 but ranging up to sixty feet, occurs at the same horizon and 

 carries Ptyctodus calceolus teeth. Besides Ptyctodus calceolus, 

 Dipterus digitatus Eastman is found near Providence. 



The sandstone probably corresponds to the Phelps sand- 

 stone of southern Missouri which, according to Weller, probably 

 represents the first deposits in the advancing Mississippian 

 sea. There is an abrupt change from the sandstone to the lime- 

 stone above, but there was no erosion between. It seems prob- 

 able to the writer that the fish remains were derived from the 

 erosion of the underlying shale and that the fishes did not live 

 in the sea in which the sandstone was deposited. Their re- 

 mains should be found in the overlying limestone if they lived 

 in the sea, and the remains found in the sandstone are always 

 greatly worn. 



The arthrodires in the fauna of the shales of Louisiana 

 are valuable in fixing the upper limit of the Devonian. The 

 Louisiana limestone of northeastern Missouri is generally con- 

 ceded as belonging to the low^er part of the Mississippian, and 

 Mr. Rowley has thought that the underlying shales should 

 be classed with the limestone. The arthrodire remains occur 

 to within an inch of the Louisiana, but no trace of them has 

 ever been found in the limestone. In Ohio the arthrodires 

 occur to the bottom of the Bedford shale, but none of them 

 has been found in the Berea above the Bedford. They will 

 probably prove of more value than the invertebrates in deter- 

 mining the upper limit of the Devonian in Ohio, if stratigraphic 



*Missouri Bureau of Geology and Mines, 2d series, Vol. Ill, p. 41. 





