ATTAINMENT OF VERY LOW TEMPERATURES. II 



Power absorbed during trial (a) 6.4 E. H. P. (b) 6.2 " 



Power absorbed in compressing the air 2.7 " 2.5 " 



Weight of air compressed per hour 14.4 kilos 14. i kilos 



Weight of air compressed per E. H. P 5.3 " 5.6 " 



Pressure 165 atm. 163 atm. 



E. H. P. required to compress the same 



quantity of air isothermally 2.23 2.18 



The energy put into the air is thtis between 20 and 25 percent more 

 than would be required for isothermal compression. The compression 

 is carried out in two stages, the ratio of the initial to the final pressure 

 being the same in each stage, so that thrusts on the two pistons were 

 balanced. The pressure in the coil and pipe connecting the two 

 cylinders was measured during one of the trials, and was found to vary 

 between thirteen and fourteen atmospheres. In a larger compressor, 

 working at a slower speed, the compression would approach much more 

 closely to the isothermal. The efficiency, which is only about 0.13 in 

 the Whitehead compressor, would also be much higher. 



The investigation of the Hampson air liquefier proved somewhat 

 troublesome as many factors had to be taken into consideration. It 

 was necessary to measure : 



The total quantity of air passing through the apparatus. 



The quantity of liquid air produced. 



The rate of absorption of heat by the coil through conduction, in 

 terms of the liquid air evaporated. 



The loss, through evaporation, of liquid air on drawing off the 

 liquid from the cock at the bottom of the apparatus. 



The pressure and temperature of the air entering and leaving the 

 apparatus. 



The temperature of the liquid air. 



The total quantity of air passing through the apparatus was meas- 

 ured by connecting the exhaust pipe of the liquefier with a gas meter. 

 This " sixty-light " meter was placed at my disposal by the Gas Light 

 and Coke Company of London ; to the manager of the Company I wish 

 to tender my thanks. 



The liquid air was drawn off at definite intervals into a weighed 

 globular vacuum vessel, silvered between the walls. The vessel was 

 half full of liquid air at the commencement of each experiment, and 

 the rate at which liquid air evaporated froin it was determined. 



To measure the quantity of air lost by evaporation of the liquid, as 

 it issued from the cock on the Hampson apparatus, the arrangement 

 shown in fig. 2, A, B, was adopted. The mouth of the vacuum vessel 

 a was fitted with a rubber stopper b, through which passed a tube c of 

 the same diameter as the thumb-screw / on the liquid air valve e of the 

 Hampson liquefier, with which it was connected by means of a piece 



