28o Mr. Barlow on the secondary deflections produced 
needle approaches those points in which its direction is at 
right angles to the line joining its pivot and the centre of the 
ball (whicli are the points of change alluded to in the pre¬ 
ceding experiments) the secondary deflections are small and 
somewhat equivocal, the precise point of change seeming to 
have a reference, not to position only, but also to the amount 
of deterioration produced in the needle. 
These curious results are important for two reasons ; first, 
as showing the necessity, in making numerical magnetic 
experiments, of being very particular as to the most perfect 
uniformity in the construction of the needles employed, as 
well as in the communication of magnetism to them ; and as 
it is probable that this uniformity can never be completely 
attained, it will enable us to account for some of those small 
irregularities, which will attend the most careful experi¬ 
ments, without attributing them to errors of observation or 
adjustment 
Secondly, these results are interesting, as amounting 
almost to a demonstration of the truth of that theory of 
magnetism, so very generally, but not universally admitted, 
viz. “ that iron becomes magnetic by induction from the 
earth.’’ 
In the first edition of my Essay, I was led by the apparent 
simplicity of the hypothesis, to adopt a particular view of 
this subject, which had been pointed out to me, and which 
referred the deflection of the needle to the simple central 
attraction of the ball on its two opposite extremities; and 
although, from various analogies and other circumstances, I 
saw reason to change |ny opinion on the subject in the second 
edition, and to adopt the induction hypothesis, yet I never 
