1880 .] 
67 
[Nichols 
about mechanically a circulation of water throughout the entire 
body of the lake, even to great depths.” 
In this same connection, I am informed by Mr. Chesbrough, 
that at the crib of the Chicago water-works, where the water, as 
already stated, is thirty-two feet deep, there is at times a difference 
of 10° F. between top and bottom, especially after a long period 
of comparatively calm weather ; but a single severe storm on the 
lake will sometimes reduce the difference to one degree or less. 
Of course, such effects as these must be very much more marked 
along the shores of a large lake than in a small pond, No doubt 
this action of the winds is an important cause of circulation. If, 
however, we compare the observations made in different ponds 
and lakes, we find that the cooling goes on so regularly and grad- 
ually that it does not seem likely that it is dependent upon such 
an irregular and, if we may use the term, accidental cause. 
Buchanan considers that after the temperature of 4° C. has 
been acquired with tolerable uniformity by the entire mass of 
water in the pond, cooling takes place more rapidly at the sides 
of the pond where the ice seems to form first. This cooler and 
consequently lighter water being unable to balance a column of 
slightly warmer water of the same height, the cooler water is 
forced outwards towards the middle of the pond while the warmer 
water tends towards the edge and thus convection currents are 
established. Now, no doubt, in the colder months, when the 
entire mass of the water is nearly of uniform density, slight causes 
produce a greater amount of circulation than when the water is, 
as it were, stratified in layers, decreasing in temperature and thus 
increasing in density from top to bottom. It is probable, also, that 
the convection currents, supposed by Buchanan, ate actually set 
in motion to some extent when the differences in density are as 
slight as in the cooling from 4° C. to say even 0°. My own obser- 
vations lead me to believe that in the case of a pond more or less 
exposed to the wind, and with a moderately steep shore, there is 
no such marked difference between the temperature of the 
water at the edge and at the same depth in the centre of the 
pond, and that, starting from 4°, the temperature of the entire 
mass of the water may fall considerably before ice forms even at 
the edge. Of course in very shallow sheltered bays or inlett the 
