STRUCTURE OF SANDSTONE DIKE 361 



the limestone along the borders, which clearly indicates a violent contact 

 between the sand and the already consolidated limestone. The suppo- 

 sition that the fissure is due to a violent disruption of the wall, probably 

 the effect of an earthquake shock, is further borne out by the numerous 

 minute faults which occur in the Waterlime in the vicinity of the fissure 

 and elsewhere. These faults never exceed in throw a fraction of an inch, 

 and are only brought out by the very perfect banding of the Waterlime. 

 From the fact that the sand which filled the fissure is not traceable 

 along the contact away from the fissure, except in a very thin interrupted 

 bed, in which shaly limestone predominates, and sand grains are not very 

 plentiful, indicates that, some considerable erosion may have occurred 

 after the filling of the fissure, for it is hardly conceivable that all the 

 sand deposited on the surface of the limestone should have been injected 

 into the fissure. It is much more likely that a considerable bed of sand 

 had accumulated above the Bullhead, part of which was forced into the 

 fissure when it was produced. The remaining sand, as well as that cover- 

 ing the limestone away from the fissure, must have been swept away 

 subsequently, only thin residual lenses remaining in hollows of the under- 

 lying rock. Indeed, it is not improbable that the Bullhead limestone 

 was formerly much thicker, and that the dike extended higher up, an 

 indeterminable thickness of limestone and dike having been removed 

 by erosion during early Devonic times. 



Devonic Rocks adjoining the Contact in Erie County 



omsk a n y representa ti ve 



Since Schuchert (1900) and Clarke (1900) have demonstrated that the 

 Helderbergian rocks, from the Coeymans (Lower Pentamerus) limestone 

 to the Kingston (Upper Shaly) beds, inclusive, are, from the nature of 

 their faunas, referable to the Lower Devonian, it is evident that the 

 Siluric section has its best representation in western New York, where 

 all its members from the Medina sandstone to the Manlius limestone 

 are developed. Not so, however, in the case of the Devonic ; practically 

 the whole of the Lower Devonian and a great part of the lower Middle 

 Devonian (following Clarke and Schuchert's classification, 1899) are un- 

 represented. The whole of the Helderbergian series of rocks is wanting, 

 as well as the Esopus and Schoharie grits. The only representation of 

 the Lower Devonian is in the thin beds of shale, shaly limestone, sand, 

 and conglomerate already noted as occupying hollows in the surface of 

 the Manlius limestone. Clarke holds that the region about western 

 New York was dry land, or at least not much below sealevel, during 



