408 CARNEGIE INSTITUTION OF WASHINGTON. 



that locality occurs \\'ithin a few minutes. The slope of the water 

 surface produced by the wind is, of course, upward to leeward. 



The slopes produced by the winds at a given point on a lake are pro- 

 portional to the 2.4 powers of the wind velocities. In other words, if a 

 wind of 10 miles per hour, from the west, produces at a given point 

 on the lake an upward slope to the eastward of x feet vertical per 

 1,000 horizontal, a west wind at 20 miles per hour will produce a slope 

 upward to the eastward of 5x feet per 1,000, and one of 50 miles per 

 hour will produce one of 48a; feet per 1,000. The 2.4 powers of 10, 

 20, and 50 are respectively 251, 1,326, and 11,954. The ratio of the 

 second to the first is 5 to 1, and of the third to the first is 48 to 1. 

 In forming a physical conception of the action, it is important to note 

 that the slope increases with increase of wind velocity much more 

 rapidly between 20 and 50 miles per hour than between 10 and 20. 

 The exponent 2.4 has been determined with a moderate degree of 

 accuracy only. The final value may prove to be 2.3 or 2.5. 



WTien a wind of uniform velocity and fixed dii-ection is blowing, 

 and is the same at all points on a lake, the slope produced at each point 

 of the lake surface is primarily a function of the depth at that point, 

 and to a limited extent of the depths at the other points and of the 

 shape of the shore. The disturbance of elevation of the water surface 

 at any point is a compUcated function of all of the depths in the lake 

 and of the location of the shore-line. A method has been developed 

 which will take this complicated function into account and enable 

 one to compute the disturbance of elevation at any given point on the 

 lake surface produced by a wind from any given direction and of unit 

 velocity. Let the numerical value of this function be called 2. 

 Then the disturbance of elevation at a given point produced by any 

 wind is 



ih'-') (2) (C) 



in which h is the wind velocity and C is an arbitrary constant to be 

 determined from the observed fluctuating values of winds and water 

 elevations. C has been determined with a moderate degree of accuracy 

 from the observations at Buffalo and Cleveland. The value of C will 

 soon be deteraiined somewhat more accurately, when the computa- 

 tions for Buffalo and Cleveland are complete. The accuracy will be 

 still further increased by similar computations based on observations 

 at Milwaukee, Mackinaw City, and Harbor Beach on Lake Michigan- 

 Huron. 



If the constants and the formula just considered were applicable 

 only to the points treated on the Great Lakes, their value would be 

 limited to use in this evaporation investigation. But the function 2 is 

 based on well-established hydraulic principles which are applicable 

 everywhere. It is primarily based on the Chezy formula, connecting 



