SOME MERCURY STANDARDS OF RESISTANCE, ETC. 1 1 5 



'Mils h;ive Keen correctly noted, then tin- two results should IM- in perfect agreement 

 within the limits of the errors of observation. ( Vrlain correct ions have, of course, to 

 be made. 



The final result of Dr. GLAZEHROOK'H investigation was that a column of mercury, 

 1 metre long and 1 sq. millim. in cross-section, had a resistance at C. equal to 

 0-95352 B.A. unit; or, to put it differently, the length of a column of mercury of 

 the same cross-section as before, and at the same temperature, having a resistance of 

 I B.A. unit, was 104'87 5 centims. The nuiiil>er OD8667 was adopted as the ratio 

 of the B.A. unit to the absolute ohm. This number is the mean of several deter- 

 minations made by Lord RAYLEIUH and Dr. GLAZEBROOK at the Cavendish Laboratory, 

 Cambridge. The length of the mercury column representing the ohm was thus 

 calculated to be 106'29 1 centims. 



In order to compare the above result with that contained in this paper, several 

 corrections have been introduced. 



Correction (1). The number adopted at the present time as representing the ratio 

 of the B.A. unit to the absolute ohm is not that employed by Dr. GLAZEBROOK in 

 1888. The history of the change is contained in the reports of the British Association 

 for 1890, 1891, and 1892. The British Association Committee in 1890 discussed the 

 results of a number of investigations dealing with the subject. The decision finally 

 arrived at was, that for practical purposes the resistance of a column of mercury, 

 106-3 centims. long, 1 sq. millim. in cross-section, and at a temperature of C., was 

 sufficiently near to the absolute ohm ; also, that the number 0'9866 best expressed the 

 ratio of the B.A. unit to the ohm. This number was adopted, and at the present 

 time is still adhered to. In order to evaluate differences, the ratio has furthermore 

 been assumed as equal to '98660. 



If this number correctly expresses the ratio, then the length of the mercury column 

 having a resistance of 1 ohm, as calculated from Dr. GLAZEBROOK'S observations, is 

 106 '29, centims. 



Correction (2). The whole of the measurements in 1888 were made by the Carey 

 Foster bridge, the mercury tubes being immersed in ice. The observations recorded 

 on p. 105 of this report show that resistances so evaluated are liable to be greater than 

 the true resistance at C. Dr. GLAZEBROOK noted the temperatures of the end 

 vessels when making comparisons ; the average temperature was 1 '3 C. A correc- 

 tion of -004 per cent, may therefore be safely applied. The length thus corrected 

 is 106'30 2 centims. 



Correction (3). Shortly previous to the observations in 1888 some change took 

 place in the standard coils already referred to. The coil " Flat " was found to have 

 fallen '01 per cent, in value relatively to the other coils. At the time the justified 

 assumption was that Flat had changed, and that the others had remained constant. 

 The value of Flat was therefore taken as '01 per cent, lower than usual (p. 355, ' Phil. 

 Trans.,' A, 1888). Intel comparisons in succeeding years rather negative this view of 



Q 2 



