SOME GLACIAL WASH-PLAINS. 81 



plains is really of constructive origin and to what extent 

 purely erosional has not been definitely determined. 



The study of creased plains becomes important in de- 

 termining chanife of water level durinof the duration of 

 the ice mass at the head of the plain, as in the case of the 

 Barrington esker-fan in Rhode Lsland, where the writer 

 has attempted to demonstrate that the water-level fell off 

 from forty to fifty feet after the construction of the delta 

 and before the disappearance of the ice at its northern 

 margin. 



Boulder-paved creases. — In those areas in which the 

 outwash of gravels took place on lower ground than that 

 on which the ice front rested, a case which occurs in the 

 Mansfield region and eastward towards Brockton, there 

 are occasionally exhibited north and south troughs, on till 

 areas, marking the outflow of water from the ice. Such 

 creases are usually paved with boulders and so resemble 

 torrent beds although the inclination of the crease may be 

 gentle. Such boulder-lines, although the material is iden- 

 tical with that of the boulder belts, should be classed with 

 the water-laid drift deposits. One or two lines of these 

 stream beds occur near North Easton on the northern bor- 

 der of the Narragansett Carboniferous area. 



Kettle-holes, ice-block holes. — Many wash-plains are in- 

 terrupted by depressions. Crateriform hollows prob- 

 ably indicate the site of buried masses of ice which on 

 melting out allowed the gravel cover to settle. A cross- 

 section of the wash-plain should here exhibit a quaquaver- 

 sal synclinal. Crosby has observed sections of this char- 

 acter near Boston. It would be an advantage to restrict 

 the term kettle-hole to depressions of this class. 



Many depressions have steep sides, with coarse detri- 

 tus, like the ice-contact phase of wash-plains in which 

 they lie. Those depressions are usually much larger than 



