OF COMPRESSED AIR. 223 



periphery of the valve. But it should be borne 

 in mind, that the issuing stream of air has to 

 overcome atmospheric resistance ; and when, 

 by diverging, it has become rarer than the 

 atmosphere against which it is acting, the 

 momentum requisite to keep it so is soon 

 expended, and the stream under the outer parts 

 of the valve, not having sufficient force to over- 

 come atmospheric resistance from without, 

 yields to it, and is brought to common 

 atmospheric density. If the velocities of the 

 stream under the different parts of the valve 

 could have been ascertained by stages of thirty- 

 second parts of an inch, in the same way that 

 the degrees of vacuum were found by the 

 heights of the mercury, it is presumed, that 

 this point would have been established by 

 experiment, instead of being left dependent on 

 an inference. 



The moving of tlie circle of greatest vacuum 

 outwards, as the valve was elevated, does, how- 

 ever, exhibit evidence of the justness of the 

 inference. When the valve was but little raised, 

 the force of the stream was expended in diverg- 

 ing a part of itself, near to the aperture; but 

 when the valve was considerably raised, the 

 superior density of the stream was not confined 

 to that part immediately over the aperture, but 



