Stevenson.] oVa [Aug. 15, 



other in a region like this, even were the ridge no greater than Brush 

 Ridge, which marks the course of the Saltsburg anticlinal. 



The accuracy of this statement will be apparent at once to one standing 

 on the National road at the summit of Chestnut Ridge, where on the west 

 he sees deep gorges made by Redstone creek, while on the east side he 

 may look down into the almost equally deep gorges made by Cheney's run. 

 The water from the former flows to the Monongahela by way of a close 

 gap through Brush Ridge ; while that from the latter, flowing against the 

 dip of the rocks, which rise southward towards the point of the Ligonier 

 canoe, passes by way of Sandy creek to the Cheat river. Each stream has a 

 rapid fall, so that no possible cutting clown of the divide by natural opera- 

 tions would suffice to divert the drainage from one side to the other. 



Altogether the more probable hj^pothesis, therefore, is that the main 

 streams flowed along or near their present lines before the axes, which they 

 cross, had been elevated so as to affect the topography ; and that the gaps 

 were eroded during the slow elevation of the folds. This process has been 

 well described by Mr. Gilbert in his Memoir on the Geology of the Henry 

 Mountains, and Maj. Powell has given instances where it is clearly impos- 

 sible that the disturbance could have ante-dated the erosion. Some of these 

 are analogous to the ones under consideration. 



The great length of time which has elapsed since this erosion began is 

 shown not only by the general wasting of the surface between the axes, 

 but also by the structure of the g;ips ; for in these, above the highest river 

 terrace, the walls slope with comparative gentleness, and in many places 

 they are deeply trenched by streams flowing directly along the axial line; 

 whereas below the line of the highest river terrace, the walls are abrupt. 

 The same conditions exist in minor gorges made by such streams as 

 George's creek, Redstone creek, Jacob's creek and Sewickley creek through 

 the axis of Brush Ridge, for there also the lower part of the gorges has 

 abrupt walls, while the upper part is wide, with walls sloping gently toward 

 the streams. It will be understood, of course, that there are occasional ex- 

 ceptions to this general statement of the conditions. 



That the erosion of the valleys ante-dated the formation of the horizontal 

 benches is shown by the distribution of the benches themselves. These 

 appear in many localities only as the crowns of flat-topped hills ; the 

 fifteenth is an extensive plain along Brush ridge, but between that and 

 Chestnut Ridge it occurs only in the crowns of hills or as the summit of 

 narrow strips stretching from one ridge to the other. Some benches are 

 seen only along the mountain slopes, while o hers line both sides of narrow 

 valleys reaching far from the rivers into the interior of the counties. All 

 of the terraces below the twelfth are shown along Pike run in 

 Washington county, though that stream breaks through the ridge 

 which shows the highest benches seen west from Chestnut Ridge. For this 

 reason, one riding from the Monongahela to Washington along the 

 National road frequently descends from higher benches to lower ones and 

 finds them all equally distinct. The lower benches can he seen along the 



