570 



NA TURE 



{April 14, 1 88 1 



In a paper on "Molecular Electro-Magnetic Induction," 

 presented to the Royal Society (March 7, 1881, I gave 

 a description of the induction currents produced by the 

 torsion of an iron wire, and the metlicd by which they are 

 rendered evident. The electro-magnetic induction-balance there 

 described is so remarkably sensitive to the slightest internal 

 strain in anywise submitted to it, that I at once perceived that 

 the instrument could not only determine any mechanical strain 

 such as torsion or longitudinal stress, but that it might indicate 

 the nature and cause of internal strains. Upon putting the 

 question to it. Does the passage of electricity through a wire 

 produce a change in its structure ? the answer came, It does, and 

 that to a very considerable extent ; for an ire n wire adjusted to 

 perfect zero, and which would remain free fnjm any strain for 

 days, becomes instantaneously changed by the first passage of a 

 current from a single cell of a Daniell's battery ; the wire has now 

 a permanent twist in a direction coinciding with that of the 

 current, which can be brought again to zero by mechanically 

 untwisting the wire, or undoing that which the passage of elec- 

 tricity has caused. Before describing the new phenomenon, I 

 will state that the only modification required in the apparatus is 

 a switch or key by means of which the telephone upon the h ire 

 circuit is thrown out of this circuit, and the current from a 

 separate battery of two bichromate cells passed through the wire 

 alone, at the same time, care being taken that no current passes 

 through the coil, but that its circuit ,-houId remain open during 

 the passage of the electric current through the wire under obser- 

 vation ; an extra switch on this circuit provides for this. The 

 reason for not allowing two cmTents to react upon each other, 

 is to avoid errors of observation which may be due to this cause 

 alone. When, however, we take an observation, the battery is 

 upon the coil and the telephone upon the wire alone. An experi- 

 ment thus consists of two operations. First all external com- 

 munications interrupted, and an electric current passed through 

 the wire ; and, second, the electric currrent taken off the wire, 

 and all ordinary communications restored. As this is done 

 rapidly by means of the sw itches, very quick observations can 

 be made, or if desired the effects of both currents can be observed 

 at the same instant. 



Now if I place upon the stress bridge a soft-iron wire 

 \ millim. diameter, 25 centims. long, I find, if no previous strain 

 existed in the wire, a perfect zero, and I can make it so either 

 by turning it slightly backwards or forwards, or by heating the 

 wire to a red heat. If I now give a torsion of this wire, I find 

 that its maximum value is with 40° torsion, and that this torsion 

 represents or produces electric currents who-e value in sonometric 

 degrees is 50 ; each degree of torsion up to 40 produces a regu- 

 lar increase, so that once knowing the value of any wire, we can 

 predict from any sonometric readings the value in torsion, or the 

 amount of torsion in the opposite direction it would require to 

 produce a perfect zero. 



If now I place this wire at zero, and thus knowing that it is 

 entirely free from strain I pass an electric current through it, I 

 find that this « ire is no longer free from strain, that it now gives 

 out induction currents of the value of 40, and although there is no 

 longer any battery current passing through this wire that the 

 strain is permanent, the outside coil neither increasing nor dimin- 

 ishing the internal strain it has received by the passage of an 

 electric current through the ware ; upon giving a'torsion to the 

 wire in one direction, I find the inductive force increase from 40 

 to 90, but in the other direction it is brought to zero, and the 

 amount of torsion, some 35°, required to bring the wire again to 

 zero represents exsctly the twist or strain that had been produced 

 instantaneously by the passage of an electric current. If I repeat 

 the experiment, but reverse the battery current sent through the 

 wire, I find an opposite twist of exactly the same value as 

 previously, and that it now requires an opposite torsion to again 

 bring the wire to zero. It is not necessary however to put on 

 an equal opposite torsion on wire to bring the currents to zero, 

 for as I have shown in my late paper, the sonometer not only 

 allov\s us to measure the force and indicate its direction, but 

 allows us to oppose an equal electric current of opposite name, 

 thus producing an electrical zero in place of the mechanical one 

 produced by torsion. 



Evidently here there has been a sudden change in the structure 

 of the wire, and it is a twist which we can both measure and 

 reproduce. The question at once becomes. Has a molar twist 

 been given to the wire such as would be detected by the arm or 

 free end of the wire, or a molecular change leaving no trace 

 upon its external form of what has passed ? 



It will be found that, notwithstanding that it requires some 40° 



of torsion to annul the effects of a passage of an electric current, 

 no visible movement nor any tendency of the free end to turn 

 in the direction of the twist it has received can be obser%'ed. I 

 believe however to have noticed a slight tremor or movement of 

 half a degree, but as I coidd not always reproduce it, and as it is 

 so -light compared with the 40° of internal twist, I have not 

 taken it into account, for if the wire is firmly fastened at both 

 ends no molar torsion being possible, except an elastic one, 

 w hich would instantly spring back to zero, the current on passing 

 produces its full effects of twist and it is permanent. Thus the 

 molecules have in some exlraordirary way rearranged themselves 

 into a permanent twist, whhout the slightest external indication 

 of so great a change having taken place. An equally remarkable 

 change takes place in aid of, or against (according to direction of 

 current) an elastic permanent strain. Thus, if I first put the wire 

 under 40° right-handed permanent torsion, I find its value to be 

 50. Now, passing the positive of battery through its free end, 

 and negative to fixed end, the induction currents rise at once in 

 value to 90; if, now, the negative is momentarily passed through 

 the free end and positive to fixed end the induced currents at 

 once fall to 10, and these effects remain, for on taking off the 

 elastic torsion the w ire no longer comes to zero, but has the full 

 twist value produced by the current. 



Tempered steel gave only one or two degrees against fifty for 

 soft iron, but supposing this might be due to its molecular 

 rigidity, I carefully brought the w ire to zero, and then observed 

 the first contact only. I found then that the first contact gave a 

 value of 40, but the second and following only one or two. By 

 bringing the wire back to zero by a momentary touch with a 

 magnet a continued force of 40, or if constant reversals were 

 used instead of a simple contact, there was constant proof of a 

 similar great molecular change by the passage of a current in 

 steel as well as iron. 



I can find no trace of the reaction of the wire upon the mag- 

 netism of the earth, as in all positions the same degree of force 

 was obtained, if great care is taken that the wire is absolutely 

 free from longitudinal magnetism ; there is however a slight 

 reaction upon its owm return wire if brought within i centim. 

 distance of the wire, and this reduces the twist some 10°. The 

 maximum effects are obtained when the return w ire is not nearer 

 than 25 centims. ; thus the action is not one produced by a 

 reaction, but by direct action upon its internal structure. 



Copper and silver wires so far show no trace of the action. I 

 believe, how ever, that a similar strain takes place in all conductors, 

 and I have obtained indirectly indications of this fact ; in order, 

 however, to verify this, would require a different method of 

 observation from the one I have described, and I have not yet 

 perfected the aj-parafus required. 



It seemed probable that if I approached a strong permanent 

 magnet to the wire, I should perceive a twist similar to that pro- 

 duced by the passage of a current ; but no such effects were 

 observed. But it has a most remarkable effect of instantly 

 bringing to zero a strain produced by the current, and, no matter 

 which pole, the effect was the same. Thus, a strain of 5o°> 

 which remains a constant, instantly disappears upon the produc- 

 tion of longitudinal magnetism, and I have found this method of 

 reducing an iron wire to zero of strain far more effective than 

 any other method yet tried, such as vibrations, heat, twisting, &c. 



It will be seen from this that the molecular arrangement set 

 up by magnetism is very different from that produced by the 

 passage of an electric cun'ent. It evidently has a structure of 

 its own, else it would not have instantly elestroyed the spiral 

 strain left by the passage of electricity if it had not taken up a 

 new form, as rendered evident in the longitudinal magnetism, 

 which we could at once perceive on the wire. This question, 

 how ever, belongs to a separate investigation, and I hope the 

 apparatus w ill aid me later in throwing some new light upon 

 this subject. 



Another method of reducing the wire to zero, after the passage 

 of a current, is to ke, p the wire in a constant state of vibration. 

 It requires in time about one minute to bring it to zero, but if, 

 on the contrary, I set the wire vibrating during the passage of 

 the current, the permanent twist becomes greater and more diffi- 

 cult to reduce to zero. 



If a wire which has internal strains is heated to redness, these 

 strains almost entirely disappear, and I can thus reduce by heat a 

 .train which a current had produced, but heat, whilst allowing of 

 greater freedom and motion of its molecules, does not prevent 

 ail internal strain being set up, for whilst heat can reduce the 

 wire to zero, after the passage of the current, the effects are 

 increased. If, during the time that the wire is at a red heat, the 



