281 
in a magnetized needle by an iron shell , &c. 
could contrive a cross experiment to decide positively be- 
tween the two theories ; for every experiment I could 
imagine, and every result I could ever obtain, which might 
be explained on one of those principles, could be as easily 
illustrated by the other. The present order of secondary 
deflections, however, is quite decisive of the point in question, 
these being all perfectly consistent with the one, and generally 
inconsistent with the other hypothesis. 
The first of the cases pointed out above, viz. when the 
needle is in the magnetic meridian of the ball, requires no 
illustration ; we may pass therefore to the second, in which 
the needle is supposed to be placed in the magnetic equator. 
Now here, on the one hypothesis, the equilibrium of the 
needle in its natural meridian is attributed to an equal and 
opposite repulsive power on its two branches, these being 
each found respectively in that hemisphere of the ball of 
the same name with itself, and are each therefore under 
repulsion. 
Consequently, if any deterioration takes place in either 
branch, that branch will be less repelled, and will therefore 
approach the ball. 
On the other hypothesis, as no repulsion is admitted, the 
equilibrium must be due to equal and opposite attractions ; 
consequently, the effect of deteriorating either branch would 
be, that that branch would recede from the ball, which is 
contrary to observation; and the same applies generally 
while the poles of the needle are in opposite hemispheres of 
the ball. 
In the third case, the explanation is nearly as simple ; for 
example : the branch of the needle nearest the centre of the 
MDCCCXXVII. O O 
