GOLDTHWAIT: SAND PLAINS OF GLACIAL LAKE SUDBURY. 271 
common, forming depressions in the top slope; near its front the plain is 
flattest, often unbroken saye for shallow depressions which lead down to 
interlobate hollows, and which seem to indicate partial scouring, subse- — 
quent to the building of the plain. Cross-sections through the body of 
the plain generally show a decrease in size of material towards the front, 
with also a decrease in thickness of the topset beds. 
The front slope, in cases where lobes are well developed, meets the 
top slope at a low obtuse angle, forming the “ brow” of the lobes. This 
point is important, because it indicates the level of the water in which 
the delta was built, the flat above the brow being almost wholly of sub- 
aerial construction and the front slope sub-aqueous. Often, however, 
the development of lobes is imperfect, or even entirely wanting, on ac- 
count of the shallowness of the lake, the irregularity of its bottom, or 
to the building forward of the plain to meet the shore or an ice-border. 
Eskers. 
Associated with these ice-bound delta deposits are usually ridges of 
gravel, eskers, which mark the courses of streams that fed the deltas. 
That these are often subglacial deposits is now recognized by most 
geologists. They concern our problem chiefly in so far as their position 
shows an area covered by the ice at the time they were formed, as their 
general direction indicates the general direction of the ice-front, being 
naturally perpendicular to it, and as their maximum height shows 
approximately the maximum height of deltas for the same stage of lake 
level. For a good description of sand plains and eskers, see Grabau 
(565-567, 574-578). 
Significance of Sand Plains. 
In what ways do sand plains, singly and collectively, throw light on 
the history of the retreat of the ice-sheet ? 
In the first place, their ice-contact borders give exact locations of the 
ice-front at the stages they represent. Taking a single sand plain, one 
can of course reconstruct only a small fragment of ice-front, and can 
hardly get a fair idea of the whole ice-dam ; but by locating and mapping 
the back slopes of all the sand plains in an extinct lake basin, the frag- 
mentary evidence may be found to fall into so systematic a scheme that 
it is possible to reconstruct with some accuracy successive lines of ice- 
front clear across the basin. The alignment of back slopes of two or 
more neighboring sand plains of the same level is not infrequently so 
