252 The Data on which we have to depend 
This difference for an area of 1,450 acres is three feet four 
inches, and will suffice to supply 90,909, at forty gallons per 
head, per day. . 
The experiment on which Mr. Hodgkinson relies to prove 
the evaporation from the surface of water, during three of 
our summer months, has many singular features. 
It was conducted on a pond on the banks of the Yarra, 
very little above the sea level, and, therefore, in the most 
favourable position to receive a lateral supply from higher 
levels. Again, decomposed trap resting on stiff clay is ex- 
ceedingly favourable to retain the winter rains from higher 
levels, and to afford a large lateral supply to a pond fifteen 
feet deep. 
It cannot be doubted that a large amount of water may be 
supplied in this way. 
In many parts of Melbourne, and particularly at the lowest 
levels, it is almost impossible to prevent the cellars being 
filled with water. And, on the Gold-fields, the difficulties 
that the diggers have to contend with from influx of water at 
low levels, and in deep excavations, is well known. 
In selecting this pond for an experiment on evaporation, 
especially when the justification of a vast expenditure of 
public money depended on the result, it was incumbent on, 
Mr. Hodgkinson to show that it contained no springs, and 
that there was no other indefinite source of supply that could 
render the experiment fallacious. 
Springs are very often found in the ponds and water-holes 
that form the beds of many of our creeks. This is a well 
ascertained fact, and was therefore deserving of careful 
consideration. 
Tn some instances the springs gush out of the rocks above 
the water line, but, in general, they are principally distingished 
by the small apparent loss from evaporation in those ponds 
in which they exist. 
The difference in this respect is very remarkable, where 
there are chains of ponds all those without springs dry up 
during the summer months, aud I have been assured by old 
colonists, and residents on the Deep Creek, and other creeks, 
that many of the ponds have*from four to six feet of waterin 
them in November, and that they dry up completely in three 
or four months. 
Nor can this be accounted for by any loss that might be 
sustained from cattle drinking at them. Where there are 
continuous chains of ponds it would be difficult to understand 
how so many should be emptied in the same manner, and, 
