283 
in a magnetized needle by an iron shell, ^c, 
common language, since in their more uniform state, they 
require the aid of the most powerful analysis to reduce them 
to determinate laws. It is sufficient for the present purpose 
to have shown, that the order of secondary deflections, dis¬ 
covered by Captain Wilson, are, in a general point of view, 
consistent with that hypothesis, which supposes the mag¬ 
netism of iron to be due to induction from the earth, and 
that they are inconsistent with that which attributes the 
deflection of a magnetised needle to the general central at¬ 
traction of the iron on its two poles or extremities, or on an 
imaginary needle passing through the pivot in the line of 
the dip. 
In adopting the hypothesis of induced magnetism in the 
Second Edition of my Essay, I only attempted the calculation 
for an indefinitely short needle, or magnetic particle. Since 
this M. Poisson has, by means of the powerful analysis he 
knows so well how to apply, obtained a general formula for 
a needle of any length ; and I have little doubt, if we pos¬ 
sessed the means of estimating the amount of deterioration, 
or the actual inequality of magnetism in the two branches of 
the needle, that all the facts I have stated would become by 
his formula a subject for calculation. 
The following experiments may perhaps in some measure 
assist towards rendering the results numerical: they were 
undertaken after the preceding part of the paper had been 
written on the suggestion of Captain Beaufort. 
Three needles were procured from Messrs. W. and T. 
Gilbert, as nearly equal in weight, length, and power as 
possible, all applicable to the same pivot and compass-box. 
The radius of each was three inches, and the number of 
