390 The American Geologist. June, 1894 
caused this water to freeze in the interstices of the rock and 
thereby to cement the rock particles firmly together. In due 
time the shale, thus frozen, was broken up by the ice, and 
fragments of it were carried southward in the till and by the 
currents of subglacial streams and were deposited, still frozen, 
in their present positions. The freezing and consequent ex- 
pansion of the water with which they were saturated must 
have loosened the already slight cohesion of the rock particles 
still more, so that, upon the melting which followed the re- 
treat of the glacier, the pebbles were left in an extraordina- 
rily friable condition. A somewhat similar instance is men- 
tioned by Prof. J. W. Spencer, who has described (Amer. 
Naturalist, Oct., 1887) the occurrence of sand boulders, which 
were broken up and transported in a frozen condition and 
then were deposited in stratified drift, thus being preserved 
in form after the ice, which had rendered this process possi- 
ble, had melted away. 
LAKE RONKONKOMA AND OTHER GLACIAL FEA- 
TURES OF LONG ISLAND. 
By John Bryson, Eastport, Long Island. 
A few weeks ago the writer paid a visit to lake Ronkon- 
koma, which is situated in the central part of Long Island, 
about fifty miles from the city of Brookl} T n. It lies in a val- 
ley between the two moraines, and is celebrated for its beauty. 
The Indians had many legends and superstitions in regard to 
it, as it was thought to have no bottom. An air of mystery 
still hangs over it, as scientists are unable to explain some of 
the phenomena connected with this wonderful little lake. 
Only a year or two ago the water began to rise without any 
apparent cause, threatening to inundate the surrounding 
country ; but before much damage was done the water began 
to subside as mysteriously as it had risen. The lake is said 
to have neither inlet nor outlet, and while this is true as to its 
present status there was certainly a time when it had several 
affluents as well as effluents. In fact, the formation of the 
lake was due to the action of subglacial streams that met at 
this point. I had long suspected this, and during my recent 
