100 



HENRY A. EOWLAND 



Table III is from a bar 2 feet long with a helix 4r| inches long near 

 one end, so that its centre was 19f inches from the end on which the 

 experiments were made, the zero-point being at the end. 



In adapting the formula to apply to the case of Table I, we may 

 assume that at the end of the bar s =o> and (7 = 0, which is equivalent 

 to assuming that the number of lines of induction which pass out at 

 the end of the rod are too small to be appreciated. 



TABLE I. 



BAR -18 INCH DIAMETER. AT END OF BAR. 



In Table II observations were not made over the whole length of 

 the rod, and the zero-point was not at the end of the bar. It is evident, 

 however, that by giving a proper value to s we may suppose the bar to 

 end at any point. As the rod is very long, expressions of the form 



Q'C" = 0'^ L C" and Q' t = rC'e-* L 

 will apply. 



In Table II the observations were near the end of the rod, and were 

 repeated several times. Neglecting the end of the rod, we have s=oo . 



In these Tables we see quite a good agreement between theory and 

 observation; but on more careful examination we observe a certain law 

 in the distribution of errors. Thus in Table I the errors of Q' are all 

 positive between and 8 inches; and this has always been found to be 

 the case at this part of the bar in all my experiments. 



The explanation of this is very simple. In obtaining the formulae,, 

 we assumed that the magnetic permeability of the bar fj. was a constant 



