murphy: notes on air resistance and pressure. 65 



velocit}' is nearly of the form Px^^a-j-bn, in which a and b are con- 

 stants and n the nnmber of fans. 



Fig. 4 shows in principle the apparatus used for finding the rel- 

 ative pressure on surfaces of different degree of curvature. 



T is a 2"8 inch tube which brings compressed air from a centrif- 

 ugal blower to the fan, O a U-shaped glass tube for measuring the 

 pressure, T a counterpoise for balancing the weight of the fan, 

 and G a weight for balancing the air pressure. 



Two sizes of fans were used, namely, 12 X24 inches, those used 

 in the experiments described above, and a 3x4 inch. 



The blast had a temperature of 78° F. and a velocity, as meas- 

 ured by an air meter, of 60 feet per second. It was directed norm- 

 ally against the fan, the end of tube being nVn inches from the fan. 



The following results were gotten: 



(i) The pressure on the concave side of y was the same as on 

 the concave side of z. 



(2) The pressure on the convex side of y was the same as on the 

 convex side of z. 



For the large fans: 



(3) The pressure on the concave surface of y was 14 per cent, 

 greater than that on x. 



(4) The pressure on the convex surface of y was 28 per cent, 

 less than on x. 



For the small fans: 



(5) The pressure on the concave surface of y was 17 per cent, 

 greater than that on x. 



(6) The pressure on the convex surface of y was 25 per cent. 

 less than that on x. 



The reduction of pressure or suction on the back side of the fans 

 was measured by the water column. It was found to be too small 

 to measure for the plane and concave fans, and -^-^ inch or 8 per 

 cent, of the pressure on the windward side for the convex fan. 



