414 



Mr. R. T. Grlazebrook on the 



respectively. The coil 1 a is connected by suitable pieces to 

 A and D, and 1 b to D and respectively. 



The resistances of the connecting-rods should be equal. 



A battery connects B and D, a galvanometer connects A 

 and C. 



If P is equal to Q and 1 a to 1 b, then there will be a balance ; 

 if this is not the case, a balance can be found by shunting P 

 or Q with a resistance W. This resistance need not be 

 accurately known. Thus 1 b can be found in terms of 1 a. Then 

 lb+la can be placed between A and D, and 2 between D 

 and C. Hence 2 can be found and so on. In this manner 

 finally the value of the whole is known. 



In the experiments referred to the values of certain 

 standard 10-ohm coils were determined, in all cases by two 

 at least of these methods, with a view to testing the accuracy 

 of the determination. Table I. gives the result. 



Table I. 



Method I. 



Method II. 



Method III. 



Difference 



expressed ae a 



fraction of the 



whole. 



10-0036 



10 0039 





- -00003 



10-0044 



10-0041 





+ -00003 



100068 





10-0070 



-•00002 



9-9950 





99946 



+ •00004 



9-9923 





9-9921 



+•00002 



99906 





99903 



+ -00003 



9-9911 





9-9909 



+ -00002 



The first method appears to give slightly higher results 

 than the others, but the differences, with one exception, would 

 be caused by an uncertainty of o, l C, and no attempt was 

 made to read the thermometers to a greater accuracy than 

 this. 



In the case of the 100 and 1000-ohm coils Method III. was 

 used ; and also a fourth method (Method IV.), that ordinarily 

 employed in testing such coils for the B. A. Committee. 



A bridge is formed consisting of the 100-ohm coil to be 

 tested, two 10-ohm coils, and a unit ; then by shunting one 

 arm as in Method III., by a resistance which does not require 



