380 Mr. E. B. H. Wade. On a new Method of 



downwards, through, the pipettes, and upwards through the G-tubes. 

 It then spreads into the drum (thus jacketing the G-tubes), and 

 finally passes through the waste W into the condenser C. This cir- 

 culation is indicated by means of the arrows in the diagram (fig. 2). 

 The condenser leads to a Woulfe's bottle V in which the pressure may 

 be adjusted to any value. 



A mercurial thermometer, T , passes through the lid b, and its 

 bulb is near the centre of the drum. Its stem is jacketed by a 

 tube, connected through a tap £ 4 , with the steam supply in A. If 

 this tap is left wide open the steam from the generator flows almost 

 entirely through it into the jacket tube and thus into the drum, 

 avoiding the alternative circuit through the G-tubes. By regulat- 

 ing the tap £ 4 we can therefore adjust the flow through the G-tubes 

 with considerable nicety. 



T x T 2 are two platinum thernionieters, and by means of them the 

 difference of the temperatures of pure water and solution, boiling 

 under similar conditions, may be ascertained. 



If it is desired to add or remove liquid to or from the G-tubes, 

 the three-way taps t x t 2 are set so as to connect them with the reser- 

 voirs, n n, instead of with the pipettes P. When the operation is being 

 carried on under ordinary pressure, the contents of the G-tubes will 

 descend into the reservoir ; on the other hand, by forcing air into 

 the reservoir, any liquid therein will ascend into the G-tube. At 

 reduced pressure the descent takes place as before, and in order to 

 make liquid ascend it is no longer necessary to blow, but merely to 

 open the three-way tap t 5 to the atmosphere. On doing this the 

 reservoir r is placed out of connection with the low pressure in the 

 Woulfe's bottle V, and into connection with the atmospheric pressure, 

 and any liquid in the reservoir at once ascends. 



Improvements. — When the apparatus came to be used, some im- 

 provements were made as the result of experience. It was found, 

 for instance, that the drum was not able completely to prevent the 

 condensation in the G-tubes, so that the depth of liquid in them 

 tended to increase. As such changes of depth were found to be a 

 source of error, at any rate when they proceeded unequally in the 

 two G-tubes, it was necessary either to equalise or prevent them. 

 Though this could easily be done by running off liquid from time to 

 time through the three-way taps t x t 2 , such a proceeding was for many 

 reasons very undesirable. Moreover, the attention of the observer 

 was quite sufficiently occupied at the bridge and galvanometer 

 (§4). 



Iii one series of experiments, therefore, a very small symmetrically 

 shaped double ring-burner was placed with one half under each 

 G-tube, and condensation was thus exactly balanced. As this burner 

 was not used in the experiments already published (loc. cit.) it is 



