106 



HENRY A. EOWLAND 



Tables V, VI, and VII are from a bar 9 feet long and -25 inch in 

 diameter. At the centre a single layer of fine wire was wound for a 

 distance of 1 foot; and the current for magnetizing the bar was sent 

 through this. The zero-point was at the centre of this helix and at the 

 centre of the bar; so that the observations on the first 6 inches include 

 the part of the bar covered by the helix. 



The values of Q' f are the sum of four observations on each end of 

 the bar and with the current reversed. The three Tables are compar- 

 able with each other, the same arbitrary unit being used for all. 



TABLE V. 

 MAGNETIZING CURRENT -176. 



Here we see an excellent confirmation of the results deduced from 

 Table IV. In Table V, where the magnetizing force is very small, and 

 where, consequently, no part of the iron has yet reached its minimum 



1 R' 

 resistance, the value of t ~ ^ R'ap. decreases continually as the value 



of Q' decreases, as it should do. In Table VI, with a higher magnetiz- 

 ing power, which was sufficient to bring a portion of the bar to about 



the minimum resistance, we see that -5 remains nearly stationary for a 



short distance from the helix and then decreases in value. In Table 

 VII, where the bar is highly magnetized and the portion near the zero- 



