DRUM LIN, OSAR AND KAME FORMATION. 265 



Returning to the region of the superficial bowlder belts in 

 Illinois, Indiana, and Ohio, we find hillocks of the kame type 

 distributed throughout the same tracts as the bowlders, indeed, 

 practically lying beneath the bowlder belts themselves. Some 

 of these I described nine years ago in the American Journal of 

 Science in an article entitled " Hillocks of Angular Gravel and 

 Disturbed Stratification." ^ Additional evidence of the same 

 import has been since gathered. Among the materials of these 

 kame -like aggregates, it is not uncommon to find a complete 

 series of gradational forms, ranging from incorporated masses 

 or tongues of typical till of the ground -moraine type, through 

 partially modified masses and layers of half-till, half-gravel, to 

 completely assorted and stratified material, thus showing every 

 stage of the derivation from the common underlying and sur- 

 rounding till. The attrition of the material shows a like grada- 

 tion. In some portions the clayey constituents of the till have 

 been simply washed out leaving the rock fragments which show 

 almost no perceptible wear from water. In others, the rounding 

 has been more considerable, and in still others, there has been 

 a reduction to the common rounded gravelly type. Even this 

 is not usually well rounded. The less modified fragments not 

 uncommonly show glacial striation. All these variations occur 

 within the limits of a single hillock, and are often so intimately 

 associated as to compel the conviction that the gravel is but a 

 partially assorted derivative from the till of the region. In 

 some of these hills the stratification is disturbed, not as though 

 the beds had been let down by the removal of ice below, but as 

 though they had been pushed horizontally by glacial pressure. 

 The essentially local derivation of the material is demonstrated 

 by the very notable presence of rock fragments derived from the 

 formations of the neighborhood. More than half the material 

 is not infrequently made up of limestone whose origin must be 

 much nearer at hand than that of the superficial bowlder belt. 

 An analysis representative of the gravel and sand of one of 

 these kames gave as much as 70 per cent, magnesian limestone. 



'Am. Jour. Sci., vol. XXVII, May 1884. Pages 378-390. 



