MR. F. F, SMITH ON THE CONSTRUCTION OF 



The essential difference between this and the first method lies in the fact that the 

 connecting tubes and standard are no longer in contact, but are separated by an 

 interval of 3 centims. The glass connecting vessels are the same as before, 1 1 it- 

 tubes passing through rubber bungs fitting tightly into the vessels. The action 

 of mercury on rubber has to be rendered negligible, and, should the erection be a 

 permanent one, contact with rubber is most inadvisable. Under such conditions the 

 ends of the tubes may be ground similar to glass stoppers, and vessels constructed 

 accordingly. Such permanent connections have been designed and will shortly be 

 employed. The rubber bungs used were first of all boiled in a weak solution 

 of caustic soda, and afterwards well sandpapered whilst wet. This operation 

 disposed of much of the free sulphur present. After several such boilings in soda 

 and in water, they were dried and immersed for some time in hot liquid paraffin wax. 

 A final washing followed, and the bungs were dried and ready for use. The thin 

 film of wax on the surface of the rubber effectually prevented any action between 

 the latter and mercury. For example, one tube was filled and remained in the ice 

 bath for 14 days, the resistance being measured daily. No difference whatever was 

 detected in these measurements. 



The arrangement adopted for the erection of the standard is shown in the figure 

 (fig. 15). 



i 



Fig. 15. 



The theoretical resistance is now greater than before, being equal to that of the 

 standard, plus the resistance of mercury between the ends of the tube and the 

 potential points. Lord RAYLEIGH has shown* that when the stream lines diverge 

 into an infinite volume of mercury, the flange of the tube being also considered as 

 infinite in extent, that the value of /in the expressions /r, and fr 2 (see p. 81) is 

 approximately '8 2. Thus under these conditions, an additional resistance is imposed 

 on the tube which is equal to the resistance of two columns of mercury of lengths 

 0'82r 1 and 0-82r 9 and of respective cross-sections Trr, 2 and TOV, where r, and r 2 are 

 the respective radii of the terminal sections of the standard. 



Such conditions, however, did not even approximately prevail in the measurements 



* " Theory of Sound," 307 and Appendix A. 



