108 THE B^RESH- WATER LOCHS OF SCOTLAND 
of nearly uniform temperature. Below this there is a layer of water 
of rapidly varying temperature — the discontinuity layer ; and below 
that again there are the abysmal waters of the lake, also of fairly 
uniform temperature. 
After the lake has become stratified in this fashion, it is, as it were, 
divided into two compartments separated by the discontinuity layer. 
As the season progresses, there is a transference of heat from the 
upper to the lower layer by conduction, convection, and otherwise, 
with the result that, while the upper layer is falling in temperature 
by parting with its heat by way of the surface, it is also parting with 
some of its heat to the lower layer, which goes on rising in tempera- 
ture, even though the lake as a whole is losing heat. The upper 
layer of uniform temperature increases in depth, the discontinuity 
layer gi-adually sinks deeper, and the difference in temperature 
between the upper and lower layers decreases, until finally the loch is 
again of uniform temperature from top to bottom. 
This is in brief the cycle of changes which takes place in lakes of 
the tropical class without reference to the effect of winds and currents. 
In lakes of the polar class there is no room for the great differences of 
temperature found in tropical lakes — differences which may amount 
to 20° Fahr. or more. In polar lakes there is only room for a varia- 
tion of about 7° Fahr. — from freezing point to the maximum density 
point. In addition to that, at the coldest period of the year, and for 
a large part of the year, the lake is covered with ice, which prevents 
the waters being disturbed by wind, and in consequence the distribu- 
tion from top to bottom is almost wholly influenced by conduction, 
and the curve representing the change of temperature from top to 
bottom is logarithmic. In summer there is a general mixing up of 
the water by wind influences, etc., so that it becomes all of uniform 
temperature. 
In lakes of the temperate class there is a mixture of the polar and 
tropical classes. In winter and spring the lake behaves as a polar 
lake, and in summer and autumn as a tropical lake.^ 
^ An elaboration of Forel's classification was proposed by Mr Whipple, 
Director of Mount Prospect Laboratory (" Classification of Lakes according to 
Temperature," American Naturalist, 1898, vol. xxxii. p. 25). He divided each of the 
three classes into three orders according as the bottom temperatures (1) are 
practically constant at or near the point of maximum density, (2) undergo annual 
fluctuations but are never very far from the point of maximum density, (3) are 
seldom very far from the surface temperatures. This subdivision was thought to 
be important from the point of view of the periods of so-called stagnation and 
circulation in lakes ; but unless the surface of a lake is protected from the action of 
the wind by a covering of ice, there is no stagnation period, and the subdivision 
does not appear to be of great interest. (See remarks on circulation of lakes, 
p. 121.) 
