158 Dr. E. Taylor Jones on 



in which the first term is generally the greatest, is in the 

 present arrangement, according to Maxwell's theory, pro- 

 portional to the induction B. In the former experiments B 

 was proportional, neglecting small corrections, simply to n/P. 



The integral I Hdl is clearly equal to the area enclosed 

 m Jo m 

 by the magnetization curve, the axis of I, and the line parallel 

 to the H-axis at distance I from it*. This area is very dif- 

 ferent in the cases of hard steel and soft iron. For this 

 reason, and in order to make the test as general as possible, 

 two bars of glass-hard steel were also experimented on. 



Preliminary Experiments. 



The ballistic galvanometer was standardized in the usual 

 way by means of a long solenoid and a secondary coil. 



The core of the solenoid consisted of a glass tube covered 

 with ebonite f; this was turned accurately cylindrical, and 

 its diameter measured with a Zeiss " thickness measurer," 

 both before and after it was wound with one layer of silk- 

 covered copper wire. From these measurements the field 

 inside the solenoid and the area of the solenoid were calcu- 

 lated in the usual way, The current in the solenoid was 

 measured by a Kelvin " platform "-galvanometer, which was 

 moreover standardized by a copper voltameter J. 



From the throw of the galvanometer-needle on reversing 

 the primary current, the ballistic sensitiveness of the galva- 

 nometer was calculated. This was not constant but varied, 

 firstly somewhat with the time and temperature, secondly 

 with the resistance of the galvanometer circuit. The first 

 changes were observed and corrected for by withdrawing a 

 small coil from a very constant permanent magnet before and 

 after each set of observations. 



This was then repeated with various resistances in the 

 galvanometer circuit. At the same time the damping was 

 also observed, i. e. the ratio m of two successive swings 

 measured from the zero. The results of these observations 

 are contained in Table 1., in which R= total resistance of 

 galvanometer circuit, <r= product of R. into the throw (<r 

 being thus proportional to the ballistic sensitiveness of the 

 galvanometer), and 



* Du Bois, Wied. Ann. xxxv. p. 146 (1888) ; Max Weber, Wied. Ann. 

 liv. p. 35 (1895). 



t Ebeling, Reichsanst.-Bericht, Zschr. Instr.-Kvnde, xv. p. 331 (1895). 



\ The instructions given in A. Gray's ' Absolute Measurements in 

 Electricity and Magnetism/ vol. ii. part ii. p. 421, were observed. 



