Stream of Air upon a Plate. 139 



Now upon keeping this gage fixed at any distance what- 

 ever from the center of the orifice, and observing its indications 

 while the upper plate was moved about, it was found that the 

 pressure was not at all affected by such motion, unless the edge 

 of the upper plate was brought very near the point of the gage, 

 when the pressure became slightly diminished. It may be con- 

 cluded from this that ccet. par. the pressure at any point of the 

 disc at a given distance from the orifice, is not affected by in- 

 creasing or diminishing the diameter of the disc, or otherwise 

 altering its figure, and this may serve to show why the small 

 discs are blown off: for if in Fig. 4, the disc be reduced to 

 the diameter of Aa, there will be no rarefaction, and it must 

 necessarily be repelled. Again, if it be made at all greater than 

 Cc, the rarefaction will not be increased, but the condensation 

 will slightly, and therefore it will sustain a rather less weight 

 upon further increasing it. 



A good experiment in illustration of all this is one that 

 was devised by Hauksbee, as long ago as 1719. He shewed that 

 when a current of air was made to pass through a small box, 

 the air contained in the box became considerably rarefied. 

 From this and similar experiments it appears that a current 

 of air communicates its motion to the particles in its im- 

 mediate neighbourhood, and carries them along with it. In 

 our experiment, then, the first portion of air when it issues from 

 the orifice instead of dispersing itself in distinct streams, com- 

 municates its motion to the air contained previously between the 

 plates, and carries it away so that the succeeding portions are 

 compelled to fill the whole space. The air may then be con- 

 sidered as issuing in successive concentric annuli from a cylin- 

 drical aperture, whose length is the circumference of the orifice, 

 and height the distance between the plates. Now as the particles 

 in each annul us issue with a certain velocity, and in lines radiating 



S 2 



