1879.] <^)o [Stevenson. 



Comparison of the levels given in this table with those given in the pre- 

 ceding notes shows some discrepancies. The altitudes used in the tables 

 approach the means, or nearly so, of the measurements At the same time, 

 no bench, aside from No. 17, shows any material variation, and in that two 

 are confounded. That two distinct benches belong here will be seen at 

 once by examination of the notes on the National road and on Westmore- 

 land county west from Chestnut Ridge. But it would be unwise to correct 

 the error in this table, as the numbers there given have been published 

 elsewhere,* and any change in the order might prove inconvenient. 



The benches seen in the southern part of the Ligonier Valley are omit- 

 ted, because their elevation cannot be regarded as satisfactorily determined. 



The variations in level exhibited by individual benches of this series are 

 so slight that they may be due either to petty variations of the barometer 

 or to errors in reading it, or to the fact that, in every case, the highest 

 point on the bench was sought, so as to determine the top of the detrital 

 deposit. As that deposit, though very thin, has suffered more or less from 

 erosion, one could not, even with perfect instruments, obtain a series of 

 measurements which would tally accurately. The extreme of variation, 

 even in No. 17 where two benches are confounded, is scarcely 18 feet ; so 

 that one is fairly justified in regarding the benches as practically horizon- 

 tal and parallel. 



The deposit on these benches sometimes contains a little clay, but sand 

 greatly predominates. No rolled or polished fragments of stone occur ; and 

 such fragments as are present belong altogether to rocks found in the im- 

 mediate vicinity. Nothing shows that running water had ever passed over 

 these plains, so that if the water were in motion, its effects must have been 

 confined within a limited space. 



The benches of the lower series have no fixed level, but are precisely 

 analogous to the bottoms of the streams which flow below them. In tabu- 

 lar form this series may be given as follows : 



1. Fifth bench of Stewartstown series above tide 1050 feet. 



2. Fourth bench of same " 980 " 



3. Third bench of same " 950 " 



4. Second bench of same " 810 " 



5. First bench of same " 790 " 



The elevations given in this list are those observed at the mouth of Cheat 

 river, and are used only to show the intervals between the benches at a lo- 

 cality where the series is well exhibited. 



It has been said that no absolute altitudes can be assigned to members of 

 this series ; the lowest bench is 790 feet above tide at the mouth of Cheat 

 river, but at Pittsburgh it is only 765 feet ; the third bench is 950 at the 

 mouth of Cheat, but 920 at the mouth of Turtle creek in Allegheny county. 

 More marked variations occur on George's creek, the Youghiogheny and 

 many other streams. 



* American Journal of Science for May, 1878; and Part II of Report on the Fay- 

 ette and Westmoreland District of Pennsylvania (1877). 



