7 1 2 THE JO URNAL OF GEOLOGY. 



the bounding bluffs, scores and perhaps hundreds of feet in 

 height, may be of drift, while at adjacent points close at hand the 

 entire faces of the bluffs may be composed of rock alone, or of 

 rock no more than drift-coated. It follows that the surface of the 

 subjacent rock is sometimes very uneven, and that its irregulari- 

 ties do not always stand in definite relation to the present topog- 

 raphy. Where extensive excavations are favorably situated, it 

 may sometimes be shown that the roughnesses of the rock surface 

 beneath the drift are due to the existence of deep valleys exca- 

 vated in the rock surface before the drift was deposited, and that 

 these valleys do not now appear at the surface because they have 

 been filled by the drift, and thus obliterated as surface features. 

 Since the deposition of the drift, the rain and the rivers have 

 in many places carved out new valleys in its surface. Sometimes 

 these valleys, developed since the deposition of the drift, have 

 sunk themselves through it, and into the rock below. 



While abrupt variations in thickness characterize the drift of 

 certain regions, and especially regions of considerable relief, 

 they are not universal. Over large areas its thickness is nearly 

 uniform. This is more likely to be true in plane areas than in 

 areas of marked relief. If the thickness of the drift be approxi- 

 mately uniform in flat regions, it follows that the surface of the 

 underlying rock must be approximately level. Where the depth 

 of the drift is uniform, or nearly so, it may be scores or even 

 hundreds of feet, or it may be very slight, affording no more 

 than a thin soil and subsoil. It is even true, now and then, 

 that areas a few or many miles in extent are nearly free from 

 drift within the very heart of larger tracts, which, with these 

 exceptions, are deeply covered. While the thickness or thinness 

 of the drift is measurably independent of both altitude and topog- 

 raphy, it is rather more commonly thick in low and nearly level 

 regions than on high and rough areas. Apart from all considera- 

 tions of altitude and topography, the distribution of thick bodies of 

 drift is not altogether fortuitous. Regarding the drift sheet as a 

 unit, its greatest average thickness is neither at its extreme edge 

 nor at its centre, but somewhere between the two positions and 



