14 



and with sharp metallic clicks as the portions of the boiling mercury 

 strike each other or the walls of the tube. When it is apparent that 

 the gases on the walls of the tube have been driven off sufficiently, a 

 fresh quantity — 3 or 4 inches — of warm mercury is added, and this 

 portion then heated and boiled. The line of separation betweens the 

 new and the old mercury is rendered plainly conspicuous by the 

 frosted appearance previously mentioned. 



These operations are repeated until the mercury reaches 3 or 4 

 inches from the tip of the tube, the latter portion being filled by the 

 careful use of the funnel without boiling. If the walls of the tube 

 are clean and dry the boiling method is easy to employ and gives 

 very high vacua. 



35. Cleaning large tubes. — Tubes that have contained mercury of 

 which oxidized and impure portions may still adhere more or less 

 closely to the wall should first be treated with dilute nitric acid (1 

 part in 20), and then thoroughly rinsed with plenty of water. 

 Ammonia or some other alkali may be added, if desired, after which 

 the operations described below should be followed. 



Introduce several inches of soapy water and whiting with tissue- 

 paper pulp. It is often easiest to put into the tube several small 

 sheets of cheap straw or manila tissue paper, and add the water and 

 whiting afterwards. 



This creamy mass can be strongly shaken about inside the tube 

 and serves to scour and clean the walls in a very satisfactory manner. 

 It is then removed by copious rinsing with clean water and after- 

 wards with distilled water. After draining some minutes, strong 

 alcohol in moderate quantities is several times introduced and suc- 

 cessively drained out and the tube given a final draining for half an 

 hour or so, if convenient, after which it is ready for drying and 

 filling. 



36. Air-pwnp method of filling. — This method, with numerous 

 modifications, has been employed by the writer in a large number of 

 cases with very satisfactory results. The method requires a good air 

 pump, drying tubes, beakers, burners, stands, etc., such as are -gen- 

 erally available in any physical laboratory. 



The apparatus is arranged as shown in Figure 15. The exhaustion 

 and funnel tube F/ will probably require to be made up to suit re- 

 quirements by some one a little familiar with simple glass-blowing 

 operations. For most purposes this may be attached to the barom- 

 eter tube by a short piece of soft pure rubber tubing. The outside 

 end of the funnel is drawn down into a long capillary extension 

 which is bent several times, as shown, so as to dip into the cup of 

 mercury M. Too fine a capillary should be avoided, and it is gener- 

 ally necessary and easy to weaken the capillary, as at a, by heating it 

 a little, so that later the tube will break off at this point when a tor- 

 sional strain is put upon it by twisting the bent extremity ate. 

 The point at c is closed by fusion, to begin with. A stopcock may 

 be employed, as at d, and in this case the breaking of the tube is not 

 required, but if the stopcock leaks even a little, the result may be 

 defective and the arrangement first described is often best. 



37. Drying and filling the tube, — In order to dry the tube it is 

 alternately nearly exhausted and dry air admitted while the walls are 

 more or less continuously heated by playing over the tube with the 



