66 MEETING, MONDAY, APRIL 4. 



least resistance. Near the Atlantic coast the movement 

 tended to be at right angles to the shore, and in the valley 

 of the Mississippi the movement converged somewhat 

 towards the axis of the valley. For a time, however, 

 there was a movement up the valley of the St. Lawrence, 

 and parallel with the axis of Lake Erie. 



Upon this map of New England (10x10 ft.) is shown 

 the system of gravel ridges, or "kames," explored since I 

 described Indian Kidge in 1875. These investigations 

 have been carried on chiefly by Mr. Warren Upham of 

 the New Hampshire Geological Survey, Prof. George H. 

 Stone of Kent's Hill, Maine, and myself. It appears 

 that in passing from New Brunswick to the Connecticut 

 river, parallel to the shore, forty or more series of gravel 

 ridges are crossed. The direction of these conforms 

 closely to that of the glacial striae, and they are in the 

 main parallel with one another. While following in 

 general the direction of the valleys, they sometimes cross 

 them and pass over moderate elevations. The kames fre- 

 quently begin near the mountains at a height of 1000 or 

 2000 feet above the sea, and are continuous for 100 or 

 more miles to the coast. The material often varies some- 

 what abruptly in the same series, changing from sand to 

 coarse pebbles a foot or more in diameter, while fre- 

 quently larger boulders are embedded. Usually the sand 

 and gravel are in a reticulated belt of ridges from twenty 

 to one hundred feet high, standing at as sharp an angle as 

 the material will allow. But frequently the ridges give 

 place for a time to extensive plains of sand and gravel ; 

 and near the shore they usually end in such plains. These 

 lines of gravel deposit are also marked by numerous " ket- 

 tle holes," of which the dungeons near Marblehead are a 

 good example. 



The most probable theory concerning the formation of 



