212 MESSRS. W. B. BOUSFIELD AND W. ERIC BOUSFIELD 



tells us that the values realised for the international volt and ohm at the National 

 Physical Laboratory as from the 1st January, 1909, cannot differ from the absolute 

 values by more than 2 parts in 10,000, and hence, within the hmits of accuracy c 

 the ,xperimental work of this investigation, the values of the international volt and 

 ohm may be taken as identical with the absolute values. 



The cadmium cell (N.P.L. 5179) was tested at the National Physical Laboratory in 

 June 1908, and, after the preliminary work of this investigation was finished, it was 

 again tested in December, 1909. It was found to have fallen in value m the interval 

 by about 15 parts in 100,000. A battery of nine cadmium cells was used in our 

 actual experiments. Each of these cells was calibrated from the standard, and the 

 value so derived for the nine cells was 9'167 international volts at 16 7 C. 



In December, 1910, when most of the work was finished, the standard 1-ohm 

 resistance was again sent to the National Physical Laboratory. It was found that 

 this resistance had gone up by about 2 parts in 10,000. A re-calibration of the 

 bridge, on which the accuracy of the working standards depended, showed that the 

 change in the bridge resistances during this time was of the order of only 4 parts 



in 100,000. 



The battery of nine cadmium cells was sent to the National Physical Laboratory at 

 the same time, and its value was found to be 9'167 volts at 17 C., so that it had 

 undergone no sensible change. 



Our best thanks are due to Dr. GLAZEBROOK, F.K.S., and to his assistants, 

 Mr. F. E. SMITH and Dr. J. A. HARKER, F.R.S., for the trouble which they so kindly 

 gave to the calibration of the thermometers and electric standards, and for the 

 detailed information with reference to them which they furnished from time to time. 



5. Calibration of the Mercury Thermometer-resistances. For this purpose a large 

 bridge was used which, as subsequently appears, was in good working order 

 throughout the range of comparison coils used (from 10,000 ohms .to O'l ohm). By 

 means of this bridge resistances could be obtained in " bridge ohms," and in order to 

 eliminate errors in the ratio coils, and to obtain results in international ohms, 

 corrective factors were obtained at different temperatures for resistances of the 

 approximate value of the resistances which were being calibrated. Thus, in measuring 

 resistances of the value of 1 to 2 ohms, the corrective factor for reducing bridge ohms 

 to international ohms was found to vary from 0'99985, at a temperature of 10 C., to 

 1 '00000 at 19 C. In measuring resistances of 9 to 10 ohms it was found to vary 

 from 1 '00010 at 11 C. to T00021 at 18 C. 



To obtain these factors the large 10-ohm manganin standard, to which reference is 

 subsequently made, which contains coils of 1, 2, 2 and 5 ohms, was calibrated from 

 the 1-ohm primary standard by the method of substitution, and the value of each 

 coil at 20 C. was thus found in international ohms within 1 or 2 parts in 100,000 

 by reference to the primary standard. Measurements of these coils on the bridge at 

 various bridge temperatures gave the required corrective factors. 



