626 



SCIENCE PROGRESS 



terms of the vapour pressure of water at the temperature of 

 experiment. 



Recent investigation by the Earl of Berkeley and Mr. Hartley 1 

 has shown that the accuracy of the method just described may 

 be considerably improved by passing the current of air over 

 instead of through the liquids in the absorption vessels. In the 

 bubbling method there is a gradual fall of hydrostatic pressure 

 throughout the train of vessels, and experiment shows that if 

 a current of air is bubbled successively through two absorption 

 tubes containing water, there is invariably a loss of weight in the 

 second tube, even although the air is saturated on leaving the 

 first tube ; the bubbles increase in volume and take up an 

 additional quantity of water in the second tube. It may also 

 be shown that in the bubbling method particles of fine spray 

 are given off and cause loss of weight. Both these sources of 

 error are avoided by passing the air over the liquids, and pro- 

 viding that the exposed surface is large and constantly renewed. 



CD G 



Fig. i. 



Fig. 2 



Each of the special absorption vessels devised by Lord Berkeley 

 and Mr. Hartley to fulfil these conditions consists of four parallel 

 and horizontal glass tubes, a, b, c, and d, sealed together by five 

 inverted U-tubes, e. Of these five U-tubes, two make connection 

 between the adjoining ends of a and b, two make a similar 

 connection for c and d, while the fifth connects the middle of 

 b to the middle of c. The air enters and leaves the vessel by 

 two vertical glass tubes, f and g, sealed to the middle of a and 

 the middle of d respectively. Fig. i is a plan of such an 

 absorption vessel, figs. 2 and 3 are side and end elevations 

 respectively. Each vessel is filled to the extent of about one- 

 third with the required liquid, and is so fitted on a platform in 

 a constant temperature bath that it can be slowly oscillated, 

 and its two ends raised and lowered alternately. The con- 



1 Proceedings of Royal Society, 1906, A, 77, 156. 



