1883.] Theory of Magnetism based on New Researches. 187 



upon the coil ; and they have the distinctive feature that the maxi- 

 mum effect is obtained with an extremely feeble torsion, differing 

 entirely from induction obtained through the medium of the spirality 

 or increased parallelism, of the conducting wire in relation to the 

 secondary coil, as the effect here is a gradually increasing force, and 

 exceedingly feeble at the degree of torsion sufficient to produce the 

 maximum of molecular rotation. Thus we have a distinct phe- 

 nomenon of molecular induction, its law of rapid maximum separating 

 it completely from that obtained from non-magnetic metals, or from 

 spirality of electric currents. If we compare the force obtained by 

 molecular rotation with mere spirality of the electric current, we find 

 that of a single turn gives the maximum for iron, whilst for a 

 •conducting copper wire it would require a spiral of similar diameter 

 of fifty whole turns to equal or balance the power obtained from mole- 

 cular induction, thus the effect is 900 times greater for the same 

 degree of spirality ; and if we neglect its distinctive feature of rapid 

 maximum and compare it with ordinary electro-magnetic force, we 

 find that it requires upon an iron core fifteen whole turns of similar 

 •conducting insulated copper wires to produce the same force ; or for 

 the same degree of spirality of conducting wires, the electro-magnetic 

 induction thus formed is 270 times weaker than that of molecular 

 induction. Thus we have not only its distinctive feature of rapid 

 maximum, but an induced current which cannot be imitated nor 

 accounted for except on the hypothesis of molecular rotation. 



It will be seen from the following diagram that the rotation takes 

 place very rapidly with the first degree of torsion, and after 20° or 

 y^- of a complete turn shows only the increased effect due to 

 continued spirality. 



It is evident from the above diagram that the molecules are 

 rotated to their maximum during the first 20°, or in reality during 

 the portion within the limits of elasticity ; from this point the 

 molecules become rigid by the strain of torsion, and they are then 

 only rotated directly as the spirality of the wire or rod. The diagram 

 only shows the effect of a right-handed torsion, the opposite torsion 

 giving equal though opposite electric currents. 



If the wire is free from strain, no increase of electric current 

 changes its zero ; nor does it while under torsion, though we may 

 then produce a confused zero, when a powerful current is used, as the 

 •electromagnetic effects from its centre free from strain superpose 

 themselves upon those due to rotation. For this reason we should 

 not employ more than one small bichromate cell, as its most powerful 

 effects are obtained with a comparatively weak primary current. 



Knowing this, we observe that, after the passage, of an extremely 

 strong current (which may be in the same or contrary direction), and 

 return to our previous feeble current, we have exactly the same 



