113 Canadian Record of Science. 
soilis small, and the underlying limestone has no filtering 
power. The water, when it reaches the rock, must either 
flow over its surface, lie in the hollows, or form streams 
running in the wide cracks. In most places the wells must 
necessarily be very shallow, mere pits for surface water, or 
they must be excavated in the rock. In either case it is evi- 
dent that more or less surface water must find its way in. 
The filtration being very imperfect, the well is seriously 
contaminated if there is any source of contamination near: 
Some deep-well owners have built cemented stone walls 
from the rock to the surface, so as to exclude, if possible, the 
surface water. But, that even this fails to prevent pollution 
is proved by the following chlorine estimations made at my 
request by Mr. F. J. Pope, M.A. The well is very care 
fully made, partly in the rock, partly in the overlying clay 
and soil. The walls from the rock upward are good 
masonry laid in cement. ‘The samples were taken about 
every second day. The weather was broken, being 
characterized by heavy rains with intervals of fine weather. 
The experiments were made during October, 1890: 
Parts of chlorine 
per million. Weather. 
52 me ws aC ae ee Rain 
53 ae ae 2 a Rain 
HSE Oller ee ac Bs ee Rain 
53.5 a 30 regi Wererene Fine 
54 as se Be Be He Heavy rain 
54 of S6 an be Fine 
5a Be vo oe ae te Heavy rain 
54.2 hs ae Ee ue Fine 
DOROmEE ae Be Ne ale Fine 
53 A are SE se Rain beginning 
54 50 sth iiedsaes ae es Heavy rain 
545 a6 oe fe te Heavy rain 
“644 4 ais 5S #2 Raining 
55 be 3 ss Be Heavy rain 
It is to be noted that the weather before these determina- 
tions were made was, on the whole, fine, so that they form 
a record of the effect of rain on the well. An inspection of 
the numbers and the weather notes shows clearly that the 
