562 
BULLETIN OF THE BUREAU OF FISHERIES. 
the sun. Such a study will warrant far more definite conclusions and will probably 
modify details of the statement given above; but the work would consume several 
years and would require the combined efforts of many observers. 
wind-distributed heat. 
The heat budgets of lakes can also be profitably compared in another way, which 
avoids the necessity of knowing the winter minimum temperature. 
The gains of heat in a lake of the temperate type may be divided in two parts, those 
below 4° and those above. So long as the temperature of the water is below 4°, the sur- 
face water becomes denser as it warms; it tends to sink and thus carry the heat into the 
deeper water. The increase in density is small and the movement, if not aided by the 
wind, would be slow; but, under the meteorological conditions of early spring, gravity 
and wind together distribute the warm surface water very rapidly through the water of 
the lake. Thus there results a rapid and uniform warming of the water. During cold 
periods the lake loses heat; but since the surface water is only slightly warmer, if at all, 
than the lower strata, and since any cooling of the surface produces an inverse stratifica- 
tion, the losses which occur during such periods and during the night are ordinarily not 
great. Thus under usual conditions a lake moves rapidly and steadily up to the tem- 
perature of 4° during the spring. 
But after this point is reached the situation wholly changes. Increase of tempera- 
ture in the water means decreased density. Gravity, so far from being an aid to dis- 
tribution, becomes an opponent; and the wind is left to do substantially the whole work 
not only without the aid of gravity, but against the resistance which gravity offers. 
This resistance is least at 4° and grows with increasing rapidity as the temperature 
rises above 4°. It follows that the action of the wind is most effective in early spring 
and becomes less efficient as the season advances and the lake warms. In April, and May 
also, the mean temperature of the air is above that of the surface of the lake, and losses 
of heat to the air are at a minimum. Thus in ordinary seasons by far the greater part of 
the heat which sun and sky furnish to the surface of the lake in April may be stored in its 
water and is easily carried to considerable depths. The amount thus stored has never 
been measured for the New York lakes; but for those of Wisconsin as much as 80 per cent 
or even more may thus be accumulated during April, and during May from 50 to 60 per 
cent of that which reaches the surface. 
As the season advances the mean temperature of the surface rises above that of the 
air. The thermal resistance of the water to mixture increases and the mean velocity of 
the wind declines. The gains of heat fall off correspondingly. Not more than 15 or 20 
per cent of the heat which falls on the surface of the lake is stored up in June, perhaps 
not more than 5 per cent in July, while in August the gains and losses of heat nearly 
balance.® 
But during much of this period of decrease in storage, the quantity of heat delivered 
to the surface of the lake is increasing, that for July being ordinarily near the maximum 
« Tlie general principles of this statement hold for all lakes. Numbers and dates will vary somewhat with the area and depth 
of the individual lake. 
