696 
SIR WILLIAM THOMSON ON THE 
always diminished the magnetization, and the effect of removing the pull increased it*. 
The magnetizing current being then stopped, and the same operation with the weights 
repeated, I found similar effects — that is, the application of a pull diminished the 
residual magnetization, and the removal of the pull increased it. But, to my surprise, 
the effect was greater in this second case when merely residual magnetism was concerned 
than in the first case when the original magnetizing force was still in action. This 
greater effect in the second case was surprising, because the whole magnetization con- 
cerned in the first case was greater than the magnetization concerned in the second case 
by the amount of the quasi-elastic magnetization, which goes and comes again every 
time the magnetizing force is removed and reapplied. 
183. The amount of the magnetizing force used may be roughly estimated by taking 
the electromotive force of each of the three cells as one volt, or 10 8 centimetre-gramme- 
second units. Thus, as the resistance in the circuit was about '69 of an ohm, the strength 
3 x 10 s 1 
of the current must have been about . 69 x 1Q g or This was distributed in 647 turns 
of a solenoid whose axis was 28 - 7 centimetres long. The whole strength of current cir- 
culating round a length bx of the solenoid was therefore X ^ bx, and the mag- 
netizing force to which the steel wire was subjected ^TyXrLip or 123. Hence, as 
remarked by Professor Maxwell, who, in reporting on this paper, first made the pre- 
ceding estimate, we may call the magnetizing force a large one, so large, in fact, that it 
probably magnetizes the wire nearly to saturation at once. For the sake of comparison 
it may be remarked that the horizontal and vertical components of the earth’s magnetic 
force at Glasgow are about T6 and - 43 ; and in first experimenting with the apparatus now 
described, I made sure that the magnetization of the wire by the vertical component of the 
earth’s magnetizing force was not essentially concerned in any of the results, by reversing 
the magnetizing current, then applying and removing weights repeatedly, then stopping 
the current, and again putting weights on and off repeatedly. The deflections of the 
galvanometer observed in this succession of operations were not sensibly different from 
those observed previously, but in reverse directions ; that is to say, the results still 
fulfilled the preceding statements. 
The fact that the effect of pull to diminish magnetization and of taking off pull to 
increase it, was found to produce a greater difference when the magnetization was solely 
residual than when it was sustained by the continued influence of the magnetizing force, 
led me to expect that the effect of making and breaking the circuit of the magnetizing 
coil and battery should be greater for the wire when pulled than when unpulled. 
Subsequent experiments proved this to be the case. 
* Since the communication of this paper to the Royal Society I have found in "Wiedemann’s ‘ Gralvanismus,’ 
§ 499, that similar results had been obtained by Mattetjcci (‘ Annales de Chimie et de Physique,’ 1858), and by 
Yielari (Pogg. ‘ Annalen,’ 1868). — [Added May 1876.] 
