474 PROCEEDINGS OF THE ACADEMY OF [1892. 



the whole is capped by from four to five feet of fine clay. Almost 

 all of the boulders occur in the gravel. 



" From this terrace the slope is gradual down to Paxton Creek, 

 on the opposite side of which a slate hill (Allison Hill) rises very 

 abruptly, and is covered on the top by the fourth terrace. 



" The fourth terrace is 130 feet above the river, 420 A. T. At 

 Walnut and Herr streets there is a very good exposure showing the 

 gravel resting upon the slate. 



" The thickness of the deposit is about 20 feet, and it is capped by 

 fine brick clay. The gravel bed, especially at its upper part, pre- 

 sents a peculiar white appearance compared with the overlying 

 clay — the line of junction between che gravel and the clay being 

 very distinctly marked. Some rounded boulders occur in it two to 

 four feet in diameter, mostly in the gravel. I found one, however, 

 in the clay. The gravel contains a few pebbles of granite and 

 gneiss, but none were found more than two or three inches in diam- 

 eter. Above this point (420 A . T.) I could find no positive evidence 

 of water action, although I carefully examined all the surrounding 

 hills and mountains north of Harrisburg." 



The occurrence of granitic pebbles in these terraces is of great 

 significance, since it fixes them as contemporaneous with, or subse- 

 quent to, the glacial period, for there is no outcrop of this material 

 anywhere in the watershed of the Susquehanna above Harrisburg. 

 The only way the granite could have come wthin reach of Susque- 

 hanna floods was by ice transportation into its headwaters, from 

 Canada or the Adirondacks. This upper terrace, therefore, corres- 

 ponds with Mr. McGee's Columbia formation. 



In following up the river I went above the first three ranges of 

 mountains to Dauphin, and drove over Fourth Mountain to Halifax. 

 Near the mouths of the small valleys occupied by Clark and 

 Armstrong creeks, there are terrace accumulations up to about 150 

 feet above the river, composed for the most part of rounded material, 

 which might have been brought down the creeks. But at higher 

 levels there were no terraces. This was a drive up the river of 

 about fifteen miles, and took us gradually over two slopes, reaching 

 about 900 feet above the river. 



We next went up the river to Selinsgrove, about six miles below 

 Sunbury, and examined the country for a few miles east of the 

 river. Here we found the rounded pebbles of the terrace deposit 

 ceasing abruptly at a height of about 200 feet above the river, or 

 about 650 feet above tide. This corroborates Professor I. C. 

 White's observations, as recorded in G 7 p. 363 of the 2d Pa. 

 Report. 



