Davis. J 
488 
[May 18 , 
stated above that its margin at other points retains a con- 
structional form. 
Neither wind nor water alone proving sufficient to account for 
the deposition of the plateau, the agency of ice may be consid- 
ered. It is, however, expressly stated that the reason for this 
consideration does not lie in its necessity. I should strongly 
disapprove of concluding that an ice-sheet had once existed in 
southern New England simply because sand plateaus cannot 
be otherwise explained. It would in such case be safer to let the 
problem remain open and search for more facts. But in the case 
in hand, the conviction that an ice-sheet once spread over our 
country was established on excellent evidence before any special 
study was made of the sand plateaus ; its agency is indeed more 
admissible than that of the sea, whose presence in postglacial 
times at a height of a hundred feet above its present level in the 
region around Boston is not yet demonstrated. 
9. ORIGIN OF SAND PLATEAUS AS DELTAS, MARGINAL TO THE DE- 
CAYING ICE- SHEET. 
During the dwindling away of the New England ice-sheet, we 
may discover conditions of a simple character that supply all the 
processes necessary for the building of a sand plateau. We have 
obstruction of drainage by the lingering remnants of the decay- 
ing ice-sheet ; we have constrained streams, escaping from the 
surface or bottom of the ice, and bearing stones, sand, and gravel 
from place to place. When such streams flowed from the ice 
into standing water, whether it was an arm of the sea or a local 
pond, sand deltas would be built, resembling in every particular, 
as far as I can draw the comparison, the sand plateaus of the 
Auburndale district. They would be coarse near the head and 
finer near the front ; their margin would consist of concave curves 
along the head, where it was built next to the decaying ice ; their 
thus layers would descend forwards, one overlying another and 
building outwards from their source at a rapid rate towards 
the front, the head rising slightly higher as the front grew further 
forward, and the upper surface thus receiving a series of coarser 
cross-bedded gravels, such as characterize the stony delta surface 
of active streams ; the free front of the delta would grow where 
the branching or distributing streams flowed for tbe time, and 
