8o 



NATURn 



\_May 22, 1879 



the circular currents. This can be perfectly balanced by 

 placing a small coin or disk of silver or copper in opposite 

 coils ; but if an iron wire or rod is placed perpendicular 

 to the coils, then increase of inductive force takes place 

 in those coils by the conduction of induced magnetism 

 from primary to secondary, and the iron can no longer be 

 balanced by silver, copper, or any non-magnetic metal. 

 The coils must be either removed farther apart, so as to 

 reduce the increased force, or balanced by an equivalent 

 amount of iron or magnetic conduction in opposite coils. 

 An interesting case of both reduction and increase of 

 force in the same pair of coils occurs if we place a disk of 

 iron, not in the centre of coils, but in the vacant space 

 between the coils. We thus reduce the force by 150°. If, 

 in addition to this, we place iron wires perpendicular and 

 in the centre, there is increase of force, and if this increase 

 is so proportioned as to be 150°, we immediately restore 

 the balance, and we have here in the same coil two 

 separate pieces of iron, each disturbing the balance and 

 giving out loud tones, but producing no effect whatever, 

 when both are introduced at the same time, complete 

 silence being the result. 



7. These coils prove what has already been long known, 

 viz., that hard steel has a far less conducting power for 

 magnetism than soft iron, although the hard steel has a 

 far higher retaining power. This instrument demonstrates 

 a point, which I have not yet seen remarked, that 

 magnetism does not in itself change the conducting 

 powers, but that it produces a molecular change of 

 structure in iron, analogous to that of tempering ; for if 

 we balance two soft iron rods against each other, the 

 balance being made perfect by the addition of fine iron 

 wires on the weakest side, we find that on strongly 

 magnetising this bar, by drawing it across a strong com- 

 pound magnet, and on replacing it in its coil, it has lost 

 30 per cent, of its conducting power ; or if, instead of 

 magnetising we make this iron red hot and plunge it in 

 cold water, the loss of conducting power will be very 

 similar — 25° to 30°. If these experiments are repeated 

 upon various degrees of iron approaching steel in 

 character, we find that as it already possesses hardness or 

 temper, it is less and less affected by magnetism, until we 

 arrive at hard cast steel, where magnetism no longer 

 produces any change in its conducting powers. From 

 this I draw the conclusion that the effect of magnetism is 

 very similar to that of temper, and shall show under the 

 effects of strain and torsion that magnetism produces this 

 temper or strain perpendicular to the lines of magnetic 

 force. 



8. The instrument shows that a remarkable change 

 takes place in the magnetic conducting power of iron 

 and steel on subjecting the wire under examination to a 

 longitudinal strain ; for if we pass an iron wire through 

 the centre of both coils, half a millimetre diameter and 

 20 centimetres or more in length, so arranged by a wind- 

 ing key that we can apply a strain to this wire, we find a 

 magnetic conducting value, unstrained, of 100, but on 

 applying a slight strain its value rapidly increases, being 

 more than double at its breaking point. If during this 

 strain we strike the wire, we hear its musical tone, and 

 no matter how much we may wind or unwind it, provided 

 we do not pass its limits of elasticity and similar wire is 

 used, that the same musical tone will invariably give the 

 same magnetic value. Thus the note A, or 435 complete 

 vibrations per second, gave always the magnetic value of 

 160, or 60 per cent, increase of power over the unstrained 

 wire. If whilst this wire is strained, giving the value 

 160, we magnetise it by drawing over it a strong com- 

 pound magnet, the note remains the same, showing no 

 difference of tension, but its magnetic value has fallen 

 80°, being now 80 instead of 160; and this wire can 

 never again be brought by strain up to its previous high 

 conducting powers. Now as we have seen that mag- 

 netism produces no change in hard tempered steel, but 



that it does so in soft iron very analogous to that of 

 temper, and as the effect of strain would be also to 

 harden the fibres by bringing them all parallel to the line 

 of mechanical strain ; and as this improves its conduct- 

 ing power, while magnetism instantly destroys all the 

 benefits of the longitudinal mechanical strain, we can 

 only draw the conclusion that magnetism produces a 

 strain analogous to temper, but contrary to that of the 

 longitudinal mechanical strain ; in other words, that the 

 magnetic strain is produced perpendicularly to its lines 

 of force. 



This view is sustained by the effects of torsion ; for if, 

 in place of straining the wire, it is twisted, instead of 

 increasing, it rapidly decreases in magnetic conductive 

 value, each turn or twist decreasing its power of conduc- 

 tion in a remarkable constant line of decrease. At eighty 

 turns of this wire there was a decrease of 65 per cent. ; 

 at eighty-five turns the wire broke, and on testing it to 

 see if magnetism had any decreasing effect on it, I found 

 that it produced no change whatever; but this twisted 

 soft iron wire had now remarkable permanent retaining 

 powers of magnetism, being superior to tempered cast 

 steel. 



Again, if we take three similar pieces of soft iron wire, 

 leave the first for comparison in its natural condition, 

 strain the second by a longitudinal strain until it is broken, 

 and twist the third by a torsion-key until it also is broken ; 

 we find on magnetising equally these three wires, and 

 allowing ten minutes' repose, that the first or untouched 

 wire has a retaining power of magnetism of 100, the 

 second only of 80, and the third, or twisted wire, of 300. 

 I hope by the light thus given soon to be able to produce a 

 magnet whose force shall be greatly in excess of what we 

 have hitherto possessed, our difficulty at present being 

 that in order to temper steel we must heat it to redness, 

 and this allows the molecules to rearrange themselves 

 contrary to the object we have in view. 



9. There is a marked difference of the rapidity of action 

 between all metals, silver having an intense rapidity of 

 action. The induced currents from hard steel or from 

 iron strongly magnetised are much more rapid than those 

 from pure soft iron ; the tones are at once recognised, the 

 iron giving out a dull, heavy, smothered tone, whilst hard 

 steel has tones exceedingly sharp. If we desire to balance 

 iron we can only balance it by a solid mass equal to the 

 iron to be balanced. No amount of fine wires of iron can 

 balance this mass, as the time of discharge of these wires 

 is much quicker than that of a larger mass of iron. Hard 

 steel, however, can be easily balanced, not only by steel 

 but by fine iron wires, and the degree of the fineness of 

 these wires required to produce a balance gives a very 

 fair estimate of the proportionate time of discharge. The 

 rapidity of discharge has no direct relation with its elec- 

 trical conductivity, for copper is much slower than zinc, 

 and they are both superior to iron. 



10. The instrument shows a marked difference in all 

 metals, if subjected to different temperatures. The value 

 is reduced in non-magnetic metals, and this we should 

 expect from the known influence of temperature on the 

 electrical conductivity ; but in the case of iron, steel, and 

 nickel (as it has already been remarked by many), the 

 contrary takes place, namely, a far higher degree of 

 magnetic conductivity. A bar of soft iron, whose value 

 at the temperature of the room, 20° C, was 160, became 

 on heating it to 200° C, 300, that is to say, its value was 

 nearly doubled. A bar of pure nickel, whose value at 20" 

 was 150, became on heating it to 200°, 320 ; thus, in the 

 case of nickel, its value for magnetic conductivity was 

 more than doubled, and at this heat it surpassed the 

 chemically pure iron at the same heat, giving a magnetic 

 value of 320 against 300 for the iron, but at the normal 

 temperature of 20" the iron had more magnetic power of 

 conduction than nickel. Heating nickel by simply 

 plunging it into boiling water increased its force from 15° 



