1880.] 461 _ [Stone. 
in level plains, or on broad slopes, or near the top of a low divide, 
where we may suppose they drained the country to the northward. 
They originate in lee of a mountain only when a drainage basin lies 
there. The larger is the drainage basin in which a kame originates, 
the larger, in general, has been the flow of water, as is shown by the 
amount and kind of material transported, etc. In this respect they 
exactly correspond with ordinary streams of water. 
The fact that the kames penetrate our hilly regions only along low 
passes makes the kames seem to have a special fondness for low and 
level regions. But kames, when crossing hills, or when near the 
height of divides, or in narrow passes, or on steep down slopes com- 
monly disappear, wholly or for the most part, while they are largest on 
level ground oron up slopes. ‘This makes the conspicuous kames ap- 
pear more frequently in level and swampy regions, while on the 
steeper and rougher ground the kames may consist of only an occa- 
sional bed of gravel or a few scattered pebbles. Obviously, however, 
a little dome of gravel or rounded pebbles in a place where no mod- 
ern stream could have deposited or rounded them, is as truly an 
organic part of the kame as those parts where it appears as a large 
ridge — providing the gravel or pebbles are found along a line of 
probable overflow. And the evidence would be complete, if, after 
tracing a broken series of water-washed materials in this way, we 
should presently come upon a well developed kame keeping the same 
general direction. It is in this way that the kame systems here de- 
scribed have been traced. All have gaps, but all lie along continu- 
ous lines of valleys or over level ground, and all are practicable lines 
for railroads. 
When following valleys the kames are sometimes found along 
the axis, but more often they zigzag somewhat, so that they now ap- 
proach one side, and now the other. Not seldom are they found zig- 
zagging along the sloping sides of a range of hills, sometimes a hun- 
dred or more feet above the valley, and in such situations, if the slope 
is steep, they may appear much like a terrace, the result of lateral 
sliding. No tendency to follow the top of a hill has yet been made 
out. In fact, as already stated, the only surface features which seem 
to invariably determine the location of kames, are drainage basins, 
and hills more than about two hundred feet high, measured on the 
north. In many instances kames cross the beds of lakes, sometimes 
for many miles, and can be traced under the water. 
