Manchester Memoirs, Vol. li. (1907), No. 14. 3 



avoid forcing over any liquid, the level reached by the 

 glycerine in the gasometer when the cylinder B is full has 

 been previously noted. The quantity of gas remaining 

 in the compression cylinder can also be estimated by 

 temporarily closing the valve g and noting the difference 

 of pressure produced by a single stroke of the pump. 



The system described above is convenient for dealing 

 with gases which ca"n only be generated at a slow rate — 

 say, ten litres or less per hour — and pressures of 3,000 lbs. 

 per square inch can easily be obtained with an ordinary 

 hand pump. 



II. Compression of Gases to very High Pressures. 



The maximum pressure for which the ordinary type 

 of two or three-stage gas-compressor can be successfully 

 built is about three or four thousand lbs. per square inch. 

 Above this limit, difficulties connected with the reduction 

 of the clearance spaces, the efficient packing of the glands, 

 pistons, etc., and the satisfactory cooling of the working 

 parts become serious and are not easily overcome. 



These troubles can be obviated by the use of hydraulic 

 compression which has been successfully used by the 

 author to compress gases up to ten or twelve thousand 

 pounds per square inch. 



In principle the system differs but little from the one 

 referred to above, but the apparatus is, of course, of 

 considerably greater strength. 



When working at these high pressures, it is con- 

 venient to eliminate the generating plant and gas- 

 ometer, and to fill the receiver B with gas previously 

 compressed and stored. The quantity of gas dealt with at 

 each operation may thus be increased a hundredfold or 

 more, and the amount of labour involved correspondingly 

 reduced. 



