>'"\n; MK 



STANDARDS OF IM-.>I> I ANGK, ETC. 



95 



Also, the measurements resulting from different fillings differ from the mean value 

 by a very small quantity. With few exceptions, this difference is less than 

 O'OOl per cent. 



Tin- recorded measurements of the tube Z were made under slightly different 

 conditions to those of the other tubes. Both glass connecting pieces of Z were 

 fractured before any determination of the resistance had been made. The connecting 

 pieces of X were substituted for them, since the terminal sections of the standard X 

 are very similar to those of Z. The resulting resistance measurements were very 

 consistent, the values obtained being as follows : 



As a check on these observations, the fractured connecting pieces of Z were 

 re-ground and used as originally intended. The cross-sections of the ends of the 

 tubes now brought into contact were but very slightly different. Possibly there 

 existed some slight difference in the shape of the sections, but certainly the connecting 

 pieces were preferable to those of X. The results of the measurements are those 

 given in Table V. It will be observed that the values are greater than those above 

 by about 0'003 per cent. The same weight cannot, however, be attached to these 

 observations as to those of the other tubes. 



Method II. of Erectiiiy the Mercury Standards. 



For the following method of erecting mercury tubes, the advantages claimed 

 are : 



( 1 ) Greater ease in manipulation and filling ; 



(2) Greater consistency in the results, and therefore a less probable error ; 



(3) Greater ease in the reproduction of resistance over short, and, more especially, 



long intervals of time. 



The personal element is also introduced to a less extent, since in this second 

 method the operations are much less exacting. 



The disadvantage lies in the uncertainty of what is known as the " End 

 Correction " ; this renders the calculation of the theoretical resistance somewhat 

 difficult. From a reprodtiction standpoint, however, the uncertainty is of no 

 ootMequenoe, 



