Delta-Plain at Andovcr, Mass. — Mills. 169 
ready assorted by water. Such a condition has been produced 
experimentally by the writer, and while furnishing no conclus- 
ive proof its evidence tended to corroborate the facts found in 
nature. Stratification formed by downsliding of material has 
none of the compactness and continuity apparent in that formed 
by the agency of an upward stream current , as exhibited in the 
true back-set beds. There is a constant tendency, in the down- 
ward gravitation of material, for the superimposed strata to 
truncate the ends of the laminae lying immediately underneath ; 
a condition that is perfectly in accord with the laws of gravity. 
The distinction in bedding is not seen by a casual glance, but 
careful investigation with freshly made and satisfactory lo- 
cated exposures seldom fails to demonstrate the fact. If on the 
northerly side of a delta-plain exposures are made by a saggital 
cut at the juncture of feeding esker and plain, and a similar 
cut is made at some distance in the ice contact phase of the 
plain, the difference in bedding and nature of stratification is 
clearly shown. [Plate XXIV, A 'and B:] 
When the delta had reached that point of maturity at which 
the ice-stream could no longer continue aggrading, it became 
a destructive agent instead of a constructive one. The stream 
had thrust upon it the necessity of channeling its own delta. 
Two of such drainage creases have left their marks across the 
Ballardvale plain. One is traceable from the south-eastern 
frontal margin well back toward the head of the plain. The 
other though larger and better developed at its mouth, on its 
southwest margin, does not appear to have eroded so far back 
toward the head of the plain, Or was subsequently filled up. 
During the entire period of delta construction the greater vel- 
ocity and stream currents probably had a decided trend to the 
southeast, for the plain here reaches its greatest extent longi- 
tudinally. The exit of these two drainage lines from the pres- 
ent plain into the former lake basin, gives a tri-lobed outline to 
the plain. 
It is evident that if the delta plain building was cotempor- 
ary with the final stages of ice-retreat, that the ice-front must 
have been in a more or less unstable condition. From the de- 
caying ice-margin would be constantly detached crumbling 
blocks or isolated bergs, some of which, rafted out over the 
plain, may have melted and dropped their heterogeneous load 
