Davis.] 
378 
[January 18 , 
basins where kames form only the barrier, and these as already 
noted (C. 4) constitute a large proportion of the ponds in New 
England. The former are seldom large enough to appear even on 
county maps, and they are often dry in the summer. 
4. During the decline of the glacial period, many large and 
small lakes stood for variable periods in front of the ice (C. 2) 
and frequently received considerable deposits of sand and gravel, 
so as to appear as sand plains when the water was drained away. 
In northern Michigan and in Wisconsin 1 there are large plains 
known as “barrens” that I believe to be of this origin : smaller 
ones are common in New England. Again, when the ice-margin 
stood in the sea, the great quantity of detritus washed from it 
by the super and subglacial streams would tend to form a gently 
sloping plain immediately in front of the ice : such a one forms 
half of Cape Cod, south of the morainal u back-bone.” 2 These 
flat surfaces are marked by abrupt hollows or bowls, with sides 
sloping five to fifteen degrees, thirty to one hundred or more feet 
in depth, and without inlet or outlet: the smaller, typical- bowls 
are oval, and but a few hundred yards in diameter ; they grade 
into much larger depressions of irregular form, with entering 
and outflowing streams and frequent kam e-like mounds that appear 
as islands in the resulting ponds. The origin of these hollows 
is probably to be found in their occupation by isolated ice-masses 
while the surrounding space was filled with drift. 3 (See also C. 3.) 
Ponds thus formed add greatly to the number included in this 
species, though in size they are insignificant. North Germany can 
probably give many examples. 
The upper terrace plains of our larger rivers contain small 
ponds that may have had a similar origin : but it has been sug- 
gested that their basins were formed by a sinking of the originally 
even surface consequent on a slow washing or squeezing out of 
the lower layers. I know of no section that will decide between 
the two explanations. 
In Eastern Massachusetts (and doubtless in many other parts of 
New England) the ponds generally occur on the line of a kame- 
1 Geol. Wise, ill, 1880, 374, 385. 
2 Upham, Amer. Nat. 1879. 
3 Peschel extends this explanation to the preservation of Lake Neuchatel and its 
neighbors ; Ueber den Ursprung der Jura- Seen, Ausland, 1868, 1005. 
