220 ON DIVERGING STREAMS 



probably was the result of some air making its 

 way under the angular pieces m. 



The cross was now raised enough to leave 

 considerable spaces for the stream to expand 

 from its previously compressed state, and to 

 become rarefied, but no greater attenuation was 

 indicated by the mercury. And thus it appeared, 

 that when there was but little space, only l-32nd 

 of an inch, under the circular valve for the air 

 to be projected into, there was an attenuation, 

 or partial vacuum, of 1 1 inches of mercury, but 

 when the cross valve was gradually raised from 

 l-32nd to the height of half an inch from the 

 seat, and when of course there was ample room 

 for expansion, not more than l-8th of an inch 

 vacuum was indicated. 



From these various phenomena it appeared 

 that the vacuum under the circular valve was 

 produced by the spreading of the air from a 

 smaller to a larger circle, immediately after it 

 left the aperture. For on the air being prevented 

 from spreading by the pieces of wood, m, Fig. 

 6, when fastened to the seat of the valve, the 

 vacuum nearly disappeared in the streams under 

 the arms of the cross valve; but by attaching 

 the angular pieces to the cross valve, and 

 suffering both to rise together^ the full vacuum 

 of 1 J reappeared as with the circular valve. 



