52 



On the Pressure of Wind on Curved Vanes. [June 4, 



1 sq. ft. area exposed normally, and hence the curve should intersect 

 the middle line at a point where the ordinate is 100. It does not do 

 so, however, partly on account of the eddy from the frame (see pre- 

 ceding paper), and partly because the two days on which the experi- 

 ments relating to this curve were made both happened to be days on 

 which the pressure was above the average. 



The slight turning up of the line near the two ends may, perhaps, 

 be due to the smaller area of projection and consequent increase of 

 pressure per unit area in those positions, or it may be due to errors 

 of observation. The curve was obtained from experiments upon one 

 plate only, and it is not unlikely that a slightly different form might 

 have resulted from the use of a larger or smaller square plate. 



Diagram showing the Kelation between the Eesistance of a Curved Plate at Perpen- 

 dicular Incidence per Unit Area of Projection and the Curvature. 



