OF COMPRESSED AIR. 217 



and it was found that one of 4} inches diameter, 

 was what may be called the neutral size over 

 an aperture of 2| diameter; as, when it was 

 balanced it would just adhere to the seat when 

 the air was admitted, but the least weight placed 

 in the scale raised it. Valves of any size smaller 

 than this did not adhere to the seat, and would 

 therefore be proper valves for such a pipe. 



A conical valve was now procured, the greatest 

 diameter of which was 6 inches on the upper 

 side, and its least diameter was 2f inches, the 

 same as the aperture, and its thickness 1 i inches. 

 This valve being fitted into a proper seat, required 

 as many ounces to raise it from its seat as the 

 flat 6-inch valve did. See Fig. 4. 



Another conical valve, whose greatest diameter 

 was the same as the flat neutral valve, 4| inches, 

 its least diameter 2|, and its thickness 3 inches, 

 was fitted like the preceding one, into a seat of 

 equal thickness with itself. This valve however 

 if less than six ounces in weight, was blown off 

 by the blast. And thus it appeared, that a conical 

 valve may be less disposed to adhere to the seat 

 than a flat valve, the diameter of the upper sides 

 of both being the same. See Fig. 6. 



A phenomenon, singular in appearance, was 

 2 E 



