﻿310 Mr. Baly and Dr. Ramsay on the Relations of Pressure, 



Torricellian vacuum which was sealed by two siphons a and b. 

 By this means the gas in the large chamber of the pump 

 suffered at most compression up to 4 millim. of mercury, so 

 that the possibility of any considerable condensation of gas 

 on the glass walls was eliminated, and each stroke of the 

 pump removed as nearly as possible the theoretical quantity 

 of gas. 



For the preliminary exhaustion of the apparatus the water- 

 pump A was connected with the top of the mercury-pump 

 through the tubes and drying apparatus «a. When this 

 preliminary exhaustion was complete, the mercury-pump was 

 brought into action. The lowering and the raising of the 

 mercury in the pump was brought about by alternately ex- 

 hausting and filling with air the reservoir C which held the 

 mercury. This could be done by hand : the exhausting by 

 connecting it with the water-pump through the tube ft , the 

 admitting air by opening the tap 7; or, if preferred, by 

 means of an automatic apparatus D through tubes 88. In 

 this apparatus a three-way tap was turned through 90° by a 

 lever, from one side of which was hung a cup fitted with a 

 curved siphon and from the other a counterpoise, water 

 being run into the cup through a flexible tube. The working 

 of this apparatus was quite constant, and* was accurately 

 timed so that if it were left working for a known time the 

 number of the strokes of the pump could easily be calculated. 

 The mercury-reservoir C was placed at such a height that 

 under atmospheric pressure the mercury would run over the 

 top siphon a, care being taken that the distance from the 

 siphon a to the T-piece c was considerably less than 30 

 inches. In the pump used it was 24 inches. 



The mercury in the two McLeod gauges E and F was 

 regulated for purposes of reading by coarse and fine adjust- 

 ment-screws Gr ; there was also a connexion with a movable 

 mercury-reservoir H. The tube I contained pentoxide of 

 phosphorus for drying purposes, and J was a vacuum-tube. 

 This last was absolutely necessary, as it otherwise would have 

 been utterly impossible to ascertain whether or not the appa- 

 ratus was full of the gas under examination. 



The gas to be examined was admitted through the tap K 

 out of the chamber L. By means of the apparatus shown in 

 the drawing any amount, large or small, of the gas could be 

 admitted. The chamber L could be exhausted by the con- 

 nexion through the tap M to the water-pump ; a wash-bottle 

 N containing mercury was interposed to prevent any diffusion 

 backwards of air or water-vapour into the gas. 



The amount of gas admitted was regulated by means of 



