1883.] Theory of Magnetism based on New Researches. 179 



The induction balance, while being one of the most simple instru- 

 ments as regards its construction, and most powerful as to its powers 

 of appreciating minute differences in the molecular construction of 

 metals, requires a lengthened previous practical knowledge of its use 

 and powers in order to obtain zero readings for each experiment. 

 Thus with it we can very easily obtain an effect, such as a marked 

 difference between two pieces of iron of similar size and chemical 

 composition; but to reduce these differences, separate and measure 

 them by the zero method, is of a peculiarly difficult nature. Not only 

 does the electromotive force vary with each piece of iron, but also the 

 time of its discharge, the time or duration of the effect being more 

 variable and more indicative of the molecular structure than the 

 electromotive force. Again, the form of the induction balance must 

 vary according to the nature of the experiment. In some cases, 

 where it is desirable not to pass an electric current through the metal, 

 four coils should be used, as in the first instrument I presented to tee 

 Royal Society in 1879.* In others we should use three, two, or but 

 one coil, as in the instrument which I shall describe in this paper. 

 In order to avoid complication I will only mention results which can 

 be easily obtained by this most simple form of apparatus, results 

 which I believe can only be attributed to the molecular nature of 

 magnetism. 



This theory has long been foreseen, and predicted in almost com- 

 plete perfection as regards the rotation of the molecules, by many 

 authors, the earliest of whose notices, and the most clearly defined as 

 being very near the results obtained by myself, will be found in the 

 remarkable work by De La Eive,t 1853, who in chapter iii, page 317, 

 under the title of " Influence of Molecular Actions upon Magnetism 

 produced by Dynamic Electricity," says : " We have seen that heat, 

 tension and mechanical actions generally facilitate magnetisation. 

 M. Matteucci has found that torsion and percussion, and mechanical 

 actions, not only facilitate the magnetisation produced upon soft iron 

 by a helix that is traversed by a powerful current, but they also 

 contribute, when the current has ceased to pass, to the destroying the 

 magnetism in a very rapid manner ; the same philosopher has likewise 

 observed that torsion, when it does not pass beyond certain limits, 

 augments the magnetisation produced upon steel needles by dis- 

 charges of the Ley den jar. 



" M. Marianini, who has made numerous and interesting researches 

 upon magnetisation, arrived at curious results upon the aptitude that 

 iron bars may acquire of becoming more easily magnetised in one 

 direction than in another." 



* " Proc. Koy. Soc," vol. 29, p. 56. 



t " A Treatise on Electricity, in Theory and Practice." By Aug. De La Eive. 

 London, 1853. 



X2 



