34 
H. G. FORDHAM—WATER-LEVEL IX 
As the effect of rainfall is in general hut slowly operative on 
deep-seated water-reservoirs, these two distinct periods are con¬ 
venient for the comparison of the water-level with the depth of 
surface-fall. 
I give also in a separate table the highest and lowest level of 
water in each year, and the mean highest and lowest, with the date. 
The highest level to which the water has attained is therefore 
50 feet, on two occasions (March 22nd, 1881, and February 27th, 
1883) , the lowest to which it has sunk, 5^- feet (December 16th, 
1884) , giving a maximum difference of 44\ feet. The individual 
year in which the change of level has been the greatest is 1882-3, 
from 10 feet to 50 feet, difference 40 feet, arise extending over 
exactly four months (October 27th to February 27th). The winter 
rise of level culminates occasionally in a very rapid upward move¬ 
ment during February, the rise being sometimes as much as 10, 
11, 12, or even 13 inches in the 24 hours, for several successive 
days. In 1879 and 1883 such spurts occurred. 
The long summer and autumn fall in level has usually no 
remarkable features. In June, 1886, however, after an exceptional 
rainfall in May (4*71 inches), there was a sharp and abnormal rise 
(May 25th to June 21st) of 8 feet. 
In 1879 the water rose all the summer, and reached its highest 
point in August, in consequence, apparently, of the abnormal 
summer rainfall (April to September, 20-37 inches). The succeeding 
autumn being very dry, the water subsequently fell rapidly, and 
reached a normal autumn level in October, 1880. But, generally 
speaking, it is the autumn and winter rainfall which goes to 
supply the underground water reservoirs, and the summer rains, 
however heavy, affect them but little. 
In the accompanying chart (Plate II) I have reduced my observa¬ 
tions to the 1st of each month, associating with them the curves 
for Barley and Therfield so far as they are contemporaneous, and I 
have added a graphic representation of the depth of rainfall for 
each month, shown on the 1st of the succeeding month. In this 
chart, which extends from April 1st, 1878, to April 1st, 1889, the 
vertical scale is 20 feet to 1 inch for the water standing in the wells, 
(0 being equivalent to the bottom of each well), and one half the 
actual depth of fall for the rainfall. The vertical lines stand for 
the 1st of each month, the thicker lines representing the 1st of April 
in each year. 
The relations between the changes of level in the three wells, 
and between those changes and the rainfall measured on the surface 
at Odsey, are thus seen at a glance. 
In general it will be noticed that the movements of the water 
are slightly later in time in the deeper wells on the higher ground 
at Therfield and Barley than at Odsey, while at the same time 
a close parallelism is maintained in the curves representing the 
changes of level in the three wells. 
The rise from November to March, and the subsequent fall, 
take place, with more or less modification from the mean curve, 
