Ricuarpson—Lines of Flow of Water in Saturated Soils. 299 
Call this K, at 7 degrees centigrade, 
then K, = K,(1 + 02817). 
Now if the plug had been parallel instead of widening, I 
would guess that a reduction to 6°5 em. long would make its 
conductance the same again. Now, the conductance of a tube 
of 1 sq. cm. cross-section in the plug is given by the last column 
of the toregoing table as 
Ae oe 
b) 
when seconds are the unit instead of minutes ; 
Ks 5 2°30 
5 = -90 x 10 ae 
whence it is found K,.;=1:3x 107 C.GS. units. 
This, however, is not a good way of measuring the porosity 
of the peat as it lies in the bog, because the structure is destroyed 
when it is pressed into the tube. A better method will be worked 
out later on. 
§ 3. Toe DirrErentiaL Hauarions. 
As has been stated, these are given by Boussinesq (1904e). 
He treats their solution in very considerable detail, but only when 
the lines of flow are nearly horizontal. In what follows the 
general equations are the same, except that capillarity is neglected, 
and the porosity supposed constant ; but they are solved for 
certain cases of steady motion in an entirely different manner, 
which can be applied when the slope of the surface is steep, and 
the lines of flow strongly curved, so that this method supplements 
Boussinesq’s. 
The particular examples dealt with in this paper could not 
have been worked out by Boussinesq’s method, with the possible 
exception of fig. 3. But for all that concerns nearly horizontal 
flow the reader cannot do better than refer to Boussinesq’s papers, 
a list of which is given at the end. 
After the experiment with the glass tube described above, it is 
fairly obvious that at any point in the peat 
i I f 
velocity ae ofall imimretsedi forse bodily force + slope o 
porosity pressure, 
