L. A. B. WADE. LXXVII, 
would hold 14 hundred million gallons, the second only 12 
hundred million gallons, and as the water was required for 
town supply, it would, at first glance, seem that the larger 
reservoir was the more desirable of the two. Another 
view of the case is that the evaporation alone would dry 
up the reservoirs in 6 and 12 years respectively. Fortu- 
nately, the records of an existing reservoir in the neigh- 
bourhood were available, and these shewed an annual loss 
of 74 feet by evaporation during the past 10 years. Apply- 
ing these figures to the two cases revealed the fact that if 
the larger reservoir had been in existence it would have 
failed three times, and would have been ina low condition 
for 3 years out of the 10, but if the smaller one had been 
in existence it would have overflowed three times, and 
would have been from half-full to overflowing during 8 
years out of the ten. The losses by evaporation in this 
case were much larger than the town supply, so that the 
ultimate choice fell on the deep reservoir in preference to 
the more capacious one. 
Another important question which can only be settled 
by actual experience, is that of the effectiveness of a water- 
shed. It has been the custom, hitherto, to take the results 
of Kuropean experience, and consequently some mistakes 
have been made. It is only quite recently that the Depart- 
ment has been able to collect any systematic information 
on the subject, but what has been acquired shews that 
further research will not be a waste of time. The following 
is a typical case :— 
A concrete dam was built to form a storage reservoir in 
a fairly humid district. The annual rainfall was 27 inches 
and the catchment area was 1,500 acres. As this area 
was covered with slate and possessed steep slopes, it was 
considered fairly impermeable, and no difficulty in filling 
the 40 million gallon reservoir was anticipated. The annual 
