A LIMNOLOGICAIv STUDY OF THE FINGER LAKES. 
553 
modify the character of the thermocline. This means that as soon as the thermal 
resistance to mixture is strongly felt the work of the wind is rapidly cut off, more slowly 
in the larger lake, but not at all in proportion to its increased size. 
The last column in table iii under the head of thermocline gives the temperature 
at the bottom of that region. It will be noted that these temperatures differ and show 
no relation to the temperature of the ground water. 
Hypolimnion . — In the hypolimnion the temperature falls at first rapidly; then 
more slowly, the curve approaching a straight line. The lower part of the very deep 
lakes may have a temperature nearly, or quite, the same through a considerable thickness 
of water. The division between thermocline and hypolimnion is not very definitely 
marked and is variable. The division of heat between these two regions is correspond- 
ingly uncertain. 
If similar climatic conditions are assumed, the temperature at the bottom of the 
lake varies with two factors, the size and the depth of the lake. On the size of the lake 
depends the efficiency of the wind until a certain area has been reached. This we have 
placed at the length of about lo kilometers for lakes with a mean depth of 30 meters or 
more. Six of the Finger Lakes reach or exceed this area and depth, and therefore have 
the maximum bottom temperatures possible under the conditions of the season of obser- 
vation. Two of the lakes, Canadice and Otisco, are both too small and too shallow to 
permit the wind to have the maximum effect; and two others, Conesus and Hemlock, 
are too shallow. 
In the six larger lakes the bottom temperature in general follows the depth of the 
lakes, the shallower lakes having a higher temperature. In 1910 it varied from 7.0° in 
Owasco Lake to 4.2° in Seneca Lake; and in 1911 from 5.3° to slightly above 4° in the 
same lakes. Owasco and Keuka Lakes have nearly the same maximum depth, but the 
bottom temperature of Keuka Lake is decidedly lower than that of Owasco in spite of 
its much greater length. This is due to the same cause that produced the thin epilimnion 
in Keuka Lake (p. 551). Skaneateles and Canandaigua Lakes, whieh have substantially 
the same length and depth, have also closely similar bottom temperatures, while the two 
larger and deeper lakes, Cayuga and Seneca, foUow in the order of their depth. 
In 1911 all of the bottom temperatures were lower than in 1910. The difference 
was almost the same in Owasco and Keuka Lakes (1.7° and 1.6°, respectively), and the 
same is true for Skaneateles and Canandaigua Lakes ( i .0° and 1.1°, respectively) . Cayuga 
Lake was about 0.3° lower in 1911 and Seneca Lake was between 0.1° and 0.2° lower. 
In 1 91 1 the temperature of Seneca Lake below 100 meters was very little above 4°. 
The mercury was slightly above the mark, but the reading would be less than 4.05°. 
Our deep-sea thermometer was not provided with an accessory thermometer for giving 
the temperature of the mercury at the time of reading and so making correction for the 
expansion of the mercury in the tube. It is therefore not improbable that the true 
temperature of the bottom water of Seneca Lake in 191 1 was slightly below 4.0°. 
The widely different temperatures of the hypolimnion in 1910 and 1911 undoubtedly 
reflect the difference of the weather in the spring of those years, though there are no 
