/ozt'a Carboiiifcroits Foniiatioiis — Sardeson. 311 
tions which are seen outcropping at Humboldt and Rutland, 
lying flatly under the glacial boulder clay ; that before the clay 
was put there by glaciers the surface of the county was upon 
those sandstone and limestone formations ; then it is evident 
that the water at that time soaking through a soil reached 
the cracks which all limey rocks possess and the humic 
acids in the water may have crowded or eaten out some 
crevices into channels, some of which formed sinks finally 
at the surface. This is evident in the underground channels 
as seen opening on the cliff at Rutland, which when judged 
by the same method as one would the crossing of two wagon 
tracks to determine which is the older, — these channels are 
older than the rivers channel and older than the glacier's work 
of erosion and deposition. These imderground channels also 
probably permeate the Kinderhook limestone throughout the 
area beneath the Drift, and drained at one time probably 
numerous sinks over the countr}', prior to the glacial changes. 
Xot only is it true that "the limestone which underlies 
the region now in question appears to be full of fissures, and 
as a result we have subterranean drainage," but evidently a 
system of preglacial large channels exists there. Evidently 
the two successive glaciers pu.shing from the north filled 
boulder clay over the sinks, though some of them have re- 
opened by gradually swallowing up a great quantity of clay. 
Where the Drift is thickest as north of the Bloody Run none 
have re-opened. The existing sinks now do not lie at the 
bottom of the surface glacial basins but anywhere seemingly 
that they happen to have been before, though probably some 
new ones are forming, by leakage from the sloughs as Mac- 
Bride says. 
In boring new sinks the conditions of the natural ones 
may be closely followed, /. t\, the well should enter the oolitic 
limestone and there strike open channels, or failing to do so, 
blasting in the well to shatter the rock facilitates the flow out 
of the drill hole to the channels. If such a well drains at all, 
each flood of water with its acids from the soil may be ex- 
pected to corrode the limestone, making the drain more ef- 
fective continually, in the manner of the natural sinks which 
are hardly choked. The failure of a well at Humboldt to 
serve as a sink in a riuarrv is due ()l)vi()uslv to the low level 
