GOLDTHWAIT: SAND PLAINS OF GLACIAL LAKE SUDBURY. 297 
border is steep, and several effluent eskers extend out from it. These 
two cases, only, have been noted in Lake Sudbury as delta deposits which _ 
may well have been built in small ice-enclosed lakelets whose outlets on 
the southern side, through tunnels, perhaps, are marked by eskers. This 
condition of things, however, is the exception. The sand plains in nearly 
every case have well developed lobate borders on the southern side, 
which were clearly built not against an ice-wall, but in open water. 
Often they have been built forward so far as to nearly overlap the north- 
ern end of an older esker. This occurs, for instance, at the western end 
of the largest South Lincoln plain. Here, as is usual elsewhere, the 
margin of the plain is not continuous up to the head of the esker ; there 
is a gap of a few hundred feet between the two. Instances have been 
noted (Woodworth) of cross-sections in sand plains which show where an 
esker abandoned by the withdrawal of the ice was subsequently buried 
by the forward extension of a sand delta. In a region where sand plains 
and eskers are abundant, this overlapping of eskers by deltas is almost 
inevitable. The mere topographical connection of the two sorts of de- 
posits, therefore, does not prove that they were contemporaneous and 
that the “effluent ” eskers mark outlets rather than inlets; unless as in 
the Walden plain, the border of the delta, near its junction with the 
esker, is an ice-contact rather than a lobate border. Aside from the two 
cases cited, there is nothing to suggest that the eskers in Lake Sudbury 
are other than feeding eskers. Frequently the eskers broaden out into 
gravel fans, at intervals, as is shown on the map. Some of these fans 
may well represent local lakelets in the much decayed ice; but I am in- 
clined to think that as a rule even the fans are true ice-front deposits, 
built by successive stages at the continually receding mouth of a sub- 
glacial tunnel. 
(4) “The elevation of all the plains of the system in this part of 
Lake Charles is very nearly or quite uniform, at about 150 feet above 
tide.” (Clapp, c, 208). Clapp’s own figures, however, place the range 
of discordance of altitudes at no less than 30 feet, —from 140 to 170 
feet. This imperfect accordance of altitudes is thought to indicate a 
state of imperfect connection between the lakelets, by super- or sub- 
glacial streams. 
Such an explanation whether correct or not, is certainly convenient. 
It might be successfully applied to any group of discordant sand _ plains, 
where the discordance amounts to even more than 30 feet. Recogniz- 
ing, therefore, that this theory would account for the differences in delta 
levels in Lake Sudbury, one is tempted to ascribe all discordance there 
