385 



will be approximately true. The quantities p, f&amp;gt; 9 v, in 

 this equation express the pressure, density, and velocity at 

 any point of the curve. Integrating from the extremity 

 of the curve in the vessel A to the orifice, we have 



f 



\J P 



Supposing the difference between P, P not too great, 

 the fluid will rush into the receiver in a stream, and there 

 will be a backward current on each side to diffuse the 

 gas over the whole vessel. Taking some section where the 

 fluid may be supposed to move with the same velocity 

 throughout, the quantity that runs past in a unit of time, 

 measured in volumes of gas at density D, is 



where p and p express the pressure and density at the sec 

 tion, and m is a numerical coefficient depending on the 

 nature of the orifice and the unit in which p is measured. 

 We are ignorant of the true values of p and p-, but if the 

 vessel be very large, and the difference of pressures P, P 

 small, the gas on each side of the stream will be nearly 

 stagnant, and we may substitute for p, p, the values of 

 P , D , the mean pressure and density in the receiver. If 

 p_p be so small that we can neglect its square, this 

 formula will become 



an expression which is in tolerable accordance with the 

 results of experiment. But if P and P are not nearly 

 equal, this is no longer even an approximation to the 



truth. 



C C 



