82 BULLETIN OF THE ESSEX INSTITUTE. 



kettle-holes and frequently are the sites of large glacial 

 lakes. Depressions of this class are typical ice-block 

 holes. 



A drainage crease sometimes starts from the ice-block 

 hole and traverses the plain ; such furrows do not origi- 

 nate in kettle-holes as defined in this paper. In the 

 kettle-hole the ice did not rise above plain level ; in the 

 ice-block hole, the ice once rose above plain level and 

 the drainage ran across the plain. 



Imperfect ice-block holes sometimes occur in the margin 

 of wash-plains as between the lobes of the Drownville 

 delta in Rhode Island. A similar phenomenon has been 

 reported by Fairchild in western New York. 



Large ice-block holes surrounded by the ice-contact are 

 to be distinguished from "unfilled areas" between suc- 

 cessive retreatal plains. Such unfilled areas will exhibit 

 the ice-contact about their southern margins and lobate 

 delta fronts about their northern border where later plain 

 building has carried sands into the de})ression. 



From the point of view of glacial geology, the occur- 

 rence of lakes in ice-block holes is an accident dependent 

 on the height of the water-plane in the surrounding 

 gravels. There are many ice-block holes of large size 

 without lakes. Such depressions exist in the Plymouth 

 area. 



Ice-block holes are sometimes grouped, as where in the 

 bottom of a large depression there are two or three isolated 

 deep holes. The accompanying map (fig. 2) of the Aga- 

 wam river area in Plymouth County, Mass., shows an 

 example of this mode of occurrence. In this case the 

 holes are occupied by water. 



Typical ice-block holes in this region seldom, if ever, 

 show ravines caused by streams eating back into the sur- 

 rounding terrace. Kettle-holes, on the contrary, as in 



