400 



Scientific Proceedings, Royal Dublin Society. 



k 



As the tube is required to withstand a very liigh internal pressure, the 

 sealed end should be fairly thick, and well rounded 

 internally. On the other hand, it is advisable to expose 

 the vapour as little as possible to the high tempera- 

 ture of the softened glass, for fear of partial decom- 

 position and formation of permanent gas. The method 

 adopted was to surround tlie glass tube, at the level of 

 the upper surface of the liquid, with cotton-wool on 

 which a little ether was poured. The liquid was thus 

 cooled, and the greater part of the vapour condensed. 

 The end of the tube was then carefully heated with 

 the smallest possible flame, so that, when softened, the 

 glass was forced inwards by the external atmospheric 

 pressure. The narrow end of the sealed tube thus took 

 internally the form of the bottom of a champagne 

 bottle. 



In order to obtain accurate results, it is absolutely 

 essential that the liquid in the tube should be not only 

 pure, but also free from air or other gas. In order 

 to ascertain whethiar the tube had been satisfactorily 

 filled, a little of the liquid was transferred, by inverting 

 and tapping the tube, to the sealed end. By holding 

 the tube in a nearly horizontal position, and tapping it 



vigorously near the sealed end, a short column of liquid was detached ; and 

 the tube behind the column of liquid was then gently warmed by holding it 

 between the fingers. The vapour-pressure thus rose a little and drove the 

 column towards the main quantity of liquid. By slightly cooling this with 

 tap-water a further small difference of pressure on the two sides of the 

 column was established ; and the column rapidly approached the larger 

 quantity of liquid. If any air or gas was present, it was collected in the 

 bubble which was left between the two quantities of liquid. In the absence 

 of air or gas the bubble was so small that, by tilting the tube and tapping 

 gently, it could be made to move along the liquid; audit gradually diminished 

 in size and finally disappeared. This test was always applied ; and when it 

 failed, as happened on two or three occasions, the filling was regarded as 

 unsatisfactory ; the end of the tube was then cut off, the liquid removed, and 

 the whole operation of filling the tube was repeated. 



The volume of liquid was measured at 0°, and with the tube surrounded 

 by running water of known temperature. The total capacity of the tube 

 was ascertained by inverting the tube, so that all the liquid flowed to the 



