2 EFFECTS OF WINDS AND OF 



of evaporation, which have been made on a small scale by using evaporation 

 pans with an area of a few square feet or a few square inches, by an investi- 

 gation made on the full scale of nature and under natural conditions. For 

 this purpose it was proposed to consider each of the Great Lakes in turn as 

 an evaporation pan and to evaluate from day to day (1) the change of content, 

 (2) the income, and (3) the outgo, including evaporation. There is hope 

 that these evaluations can be made with sufficient accuracy to segregate 

 that part of the outgo which is evaporation, and to determine the laws which 

 control the rate of evaporation. If that hope is realized, the laws so deter- 

 mined may be applied with a confidence based upon the fact that the 

 experimental evidence is on the full scale of nature and under natural 

 conditions. 



The change of content of any one of the Great Lakes from day to day is 

 measured by the change of elevation of the mean lake surface from day to 

 day. The area of the lake surface remains substantially constant. For 

 Lake Erie, for example, the area never varies by as much as 0.01 part from 

 9,968 square miles. If the elevation of the mean surface of the lake is ob- 

 served to have increased by 0.01 foot, the content of the lake is known to 

 have increased by the net addition of a sheet of water 0.01 foot thick having 

 an area of 9,968 square miles. Such a sheet contains about 2,800,000,000 

 cubic feet and is equivalent to the outflow through the Niagara River under 

 typical conditions for about 3.7 hours at 210,000 cubic feet per second. 



The elevation of the surface of each of the Great Lakes has been deter- 

 mined at a few selected points by recording gages operated continuously 

 day and night for many years by the United States Lake Survey. The most 

 important gages are at Tibbetts Point (New York), Buffalo (New York), 

 Cleveland (Ohio), Harbor Beach (Michigan), Mackinaw City (Michigan), 

 Milwaukee (Wisconsin), and Marquette (Michigan). As the investigation 

 progressed, it gradually became more clearly evident that the largest and 

 most serious errors encountered were those which arise from the fact that 

 the surface of any one of the Great Lakes at any given instant is not level 

 except by accident. The surface has a slope at every point due to the in- 

 fluence of winds and barometric pressures. Hence, when the elevation of 

 the surface of Lake Erie, for example, is observed at Buffalo on each day, it is 

 necessary, if one is to obtain daily values of the elevation of the mean surface 

 of Lake Erie, to apply a correction each day to correct for the disturbance of 

 the elevation on that day at Buffalo produced by the winds and barometric 

 pressures. 



The recording gages are so designed and operated that the rapid fluctua- 

 tions of elevation of the water surface which one sees as wind waves are 

 practically eliminated from the record. The elevation recorded by the gage 

 at any instant is with a high degree of accuracy, usually within 0.1 or even 

 within 0.01 foot, equal to the mean elevation for the preceding 10 minutes, 

 even though the wind waves may be causing the instantaneous elevation of 

 the water at the gage to vary through a range of several feet several times in 



