RESEARCHES IN' MAGNETISM. 
539 
In this simple form the direct magnetometric method, as a means of determining 
magnetic qualities in absolute measure, is not entirely free from objection, but it 
allows certain points to be investigated which the ballistic method is unable to deal 
with, notably the influence of time in magnetisation, and the comparative effects of 
sudden and gradual application or removal of magnetising force, and it is also particu¬ 
larly convenient in testing the effect of stress on magnetic susceptibility. 
§ 19. Demagnetisation by the method of reversals. —As an alternative to the plan of 
tapping out the magnetism, which is inapplicable to steel or hard iron, another plan 
has more frequently been used to remove residual magnetism when it has been desired 
to repeat the magnetisation of a given specimen without change of its physical 
state. This consists in subjecting the piece to a succession of magnetising forces 
with opposite signs and of gradually diminishing intensity. By means of the liquid 
slide described above, the current was slowly reduced to zero, while a rapid reversing 
key between the slide and the solenoid was kept in action. The earth’s vertical force 
was, as has been said, balanced by a steady current in a separate solenoid, and when 
the strength of this last was properly adjusted, the process of demagnetising by 
reversals was very successful in removing all traces of residual magnetism/'' It is 
clear that this process leaves the piece symmetrically conditioned as regards subse¬ 
quent magnetisation in either of the two longitudinal directions. The condition of 
a piece demagnetised in this way is no doubt different from that of a piece never before 
magnetised (indeed the curves of 3 and A, taken before and after the process, show 
that it is somewhat, though not very, different as regards susceptibility); but the 
neutral state produced by this process forms an exceedingly convenient starting 
point, to which we may recur over and over again, when w T e wish to determine the 
susceptibility or the retentiveness of the same piece under different conditions of 
stress, temperature, &c. 
§ 20. Cyclic Magnetisation of Wrought-iron Wires. —A few experiments will now 
be detailed in which well annealed! specimens of soft iron wire were subjected to a 
complete cycle of magnetisation by the successive application of equal and opposite 
magnetising forces. In each case the length of the specimen was so great as to 
realise, as nearly as possible, the condition of endlessness. In the following experi¬ 
ment, the results of which are given numerically below, and graphically in Plate 57, 
fig. 4, the specimen was a very soft iron wire 0’158 centim. in diameter, and 
64 centims. or 400 diameters in length; and the method used was the step-by-step 
ballistic method (§ 8). The steps by which <§ and £3 were varied were those shown 
in the table below. The magnetising force was first raised to 17'26, then reversed 
* A process similar in principle was applied some years ago by Mr. H. S. Maxim to cure magnetised 
watches. 
f The process of annealing consisted in drawing the wire through a blowpipe flame so slowly that 
each part of the length in turn became bright red and then cooled gradually as it moved away from the 
flame. For wires of less than 1 mm. diameter an ordinary spirit-lamp flame was generally used, without 
a blowpipe. 
