R. A. Hazen — Wind Velocity and Pressure. 245 



several turns the plate became somewhat steadier and showed 

 a slightly less pressure, possibly owing to a slight following 

 after of the air, produced by the motion of the plate. In order 

 to determine the uniformity of the times of revolution, a chrono- 

 graph was employed, the bar being made to record its own 

 turns. One of these records is here given : 



Chronograph Record of Revolutions. 



No. of the 





No. of the 





No. of the 





turn. 



Seconds. 



turn. 



Seconds. 



turn. 



Seconds. 



1 



4-8 



9 



3-9 



17 



39 



2 



4-5 



10 



3-9 



18 



4-0 



3 



4-4 



11 



3-9 



19 



3-9 



4 



4-3 



12 



3-8 



20 



4-0 



5 



4-1 



13 



3-9 



2] 



4-1 



6 



4-0 



14 



3-9 



22 



4-0 



7 



39 



15 



39 



23 



3-] 



8 



39 



16 



3-9 



24 



4-2 



This table shows that after the first 6 turns the motion be- 

 came uniform and continued so for at least 14 turns. After a 

 long series of observations the following formula was obtained : 



p= ('0032 + -00034c) SV 2 . 



There is a most remarkable and unexpected coincidence be- 

 tween this result and that of Borda, and this correspondence 

 seems to indicate that there could hardly be any constant 

 errors, as the apparatus was so entirely different from Borda's. 

 It will be noted that Hagen's formula was derived after the 

 total pressure had become uniform, and it may be said that 

 nearly the same result would have been found after the first 15 

 or 20 revolutions of the above apparatus. 



The later experiments with an arm of 16 feet developed the 

 fact that there was no increase of pressure per square foot with 

 the larger plates, as in the previous results, but that it was con- 

 stant for sizes from 16 to 576 square inch area. The formula 

 from these results was : 



£>=-0034 SV 2 , Rouse's experiments. 



All the results thus far, except those at Metz, were derived 

 with whirling machines, and it is of interest to compare them 

 with others made with an air current impinging upon the plate. 

 Such, probably, are those of Rouse in the 18th century. His 

 formula was as follows : 



^=•0049 SV 2 , Thibault's results. 



In these experiments plates of 1*162 and 1*531 square foot area 

 were exposed to direct wind pressure and gave the formula : 



i?=-00475 SV 2 . 



