248 H. A. Hazen — Wind Velocity mid Pressure. 



plates, and up to four miles per hour with larger plates, the 

 relation between the pressure and velocity may be expressed 

 by the formula, 



j»=-0034 SV 2 . 



In this, p = pressure in pounds : S=surface in square feet: V= 

 velocity in miles per hour. 



4th. Experiments are much needed, with larger plates than 

 two feet square, with bodies of other forms, and with high ve- 

 locities by a straight line motion. 



No attempt has been made in this paper to give a summary 

 of anything more than what appeared to be the better class of 

 results in this field. It is believed enough has been given to 

 show which results are the more reliable, and how much neces- 

 sity there is for further research. 



Experiments with an anemometer. -One of the more important 

 questions in connection with the use of Robinson's anemometer 

 is the determination of the relation between the velocity of the 

 wind and that of the cups. This relation has been assumed to 

 be about 3, i. e., three times the velocity of the cups equals 

 that of the wind. Some experiments in England with large 

 anemometers as compared with smaller, have shown that if the 

 former have a velocity of 2*5, the smaller would need a co- 

 efficient about a fifth larger to give the true velocity. The 

 earlier anemometers used in this country were rather clumsy 

 and unsatisfactory. A form was devised by Lieut. Gibbon 

 (Signal Service), after experiments with an electric register of 

 the actual revolutions, and a comparison with that form of 

 wheel register which gave the result nearest to that of the elec- 

 tric count. This form has never been tested for its coefficient, 

 so far as can be learned, hence a few observations were made 

 on the 16 foot arm. The anemometer used was nearly new, 

 having been run only 1,500 miles, was carefully oiled and in 

 most excellent condition. The cups were 4 inches in diameter, 

 and their distance from the axis 6*72 inches. The highest 

 velocity attained was 12 miles per hour. The values of the 

 coefficient in five sets were 2*92, 3*06, 2'84, 2-98, and. 2 -92, 

 giving a mean 2'94 instead of 3*00, the one now in use. It 

 should be noted that, intentionally, there was introduced a ten- 

 dency toward a too small value, for, after the cups had acquired 

 a momentum, toward the end of the experiment, the bar was 

 stopped rather abruptly, but the cups were allowed to come to 

 rest of their own accord. Experiments on this question are 

 still needed with a rectilinear motion and with much higher 

 velocities. 



"Washington, April 20, 1887. 



