394 Royal Society. 



bodies, so, conversely, the discoveries of numerous investigators 

 have shown that every electric current passing near or through a 

 substance produces a molecular change, which is rendered manifest 

 in all metals, liquid conductors, and even in the voltaic arc by the 

 development of sounds, especially if the substances are under the 

 influence of two currents at right angles to each other. In iron it 

 is conspicuously shown also by electrotorsion, a phenomenon I have 

 f ound and recently made known in a paper read before the Eoyal 

 Society. 



Numerous facts also support the conclusion that the molecular 

 changes referred to last as long as the current. De la Rive has 

 shown that a rod of iron, either transmittmg or encircled by an 

 electric current, emits, as long as the current lasts, a different sound 

 when struck ; and we know it also exhibits magnetism. The pe- 

 culiar optical properties of glass and other bodies with regard to 

 polarized light discovered by Faraday also continue as long as the 

 current. A rod of iron also remains twisted as long as it transmits 

 and is encircled by electric currents ; and in steel and iron the 

 molecular change (like magnetism) partly remains after the currents 

 cease, and enables the bar to remain twisted. 



That the peculiar molecular structure produced in bodies gene- 

 rally by the action of electric currents also possesses a definite di- 

 rection with regard to that of the current, is shown by the rigidly 

 definite direction of action of magnetized glass and many other trans- 

 parent bodies upon polarized light, also by the difference of con- 

 ductivity for heat and for electricity in a plate of iron parallel or 

 transverse to electric currents, by the stratified character of electric 

 discharges in rarefied gases and the action of electric currents 

 upon it, and especially by the phenomena of electrotorsion. In 

 the latest example an upward current produces a reverse direction 

 of twist to a downward one, and a right-handed current develops 

 an opposite torsion to a left-handed one ; and the two latter are 

 each internally different from the former. As each of these four 

 torsions is an outward manifestation of the collective result of in- 

 ternal molecular disturbance and possesses different properties, 

 these four cases prove the existence of four distinct molecular 

 movements and four corresponding directions of structure; and 

 the phenomena altogether are of the most rigidly definite character. 



As an electric current imparts a definite direction of molecular 

 structure to bodies, and as the attractions and repulsions of electric 

 wires are between the wires themselves and not between the cur- 

 rents, repulsion instead of attraction must be due to difference of 

 direction of structure produced by difference of direction of the cur- 

 rents. 



Although the Amperean theory has rendered immense service to 

 magnetic science, and agrees admirably with all the phenomena 

 of electromagnetic attraction, repulsion, and motion, it is in some 

 respects defective ; it assumes that magnetism is due to innume- 

 rable little electric currents continually circulating in one uniform 

 direction round the molecules of the iron ; but there is no known 



