42 THE UNDERGROUND WATERS OF NORTH-WEST YORKSHIRE. 
Each set of joints may be regarded as a sort of under- 
ground valley through which, with great retardation by friction, 
a stream of water flows. There will be between these two lines 
of flow an area of saturated rock receiving the direct percolation 
of rainfall from the intervening tract of the limestone plateau. 
This saturated area will drain into the subterranean " valleys " 
and produce saturation gradients following the rules already defined. 
Suppose, now, a "flush" to be sent down from Malham Tarn 
into valley No. 2 (Fig. 7). Before any effect can be produced upon 
the flow in valley No. 1 the water must accumulate underground 
to such an extent as to cause a rise of the saturation-level over 
the intervening watershed. The readiness with which this 
... VALLEY VALLEY 
W No. I. No 2. 
M 
■1 
'SEimniiii 
t 
Mnlf: 
iilillli 
Fig. 7. 
can be accomplished will depend upon two factors — (1) the 
magnitude and duration of the "flush"; (2) the height of the 
underground w-atershed. The height of the watershed is evidently 
a very material factor, for upon it will depend the volume of 
water which must accumulate in the subterranean valley before 
an overflow can take place. The magnitude of the flush will be, 
for obvious reasons, of equal importance. When the saturation 
level is high a large volume of water will be needed to overflow 
it; when the level is low a smaller quantity will suffice. 
Mr. Tate's experiments were made in a season of normal 
rainfall, but there appears to have been a very large volume of 
water sent down, for both sluices of the Tarn were opened, and 
it is stated that on the first day the sinks were overflowed in 
