July 29, 1880] 



NATURE 



505 



of the transversal vibrations is much greater than with the thick. 

 Using a brass wire 15 cm. long and '4 mm. in diameter, I ob- 

 tained with a tolerably powerful horse-shoe magnet transversal 

 vibrations of 2 mm. amplitude or more. 



The wire telephone, when used in this way, is pretty sensitive 

 to magnetic influences. The pre-ence of the pole of a bar 

 magnet could be detected at a distance of several inches from 

 the wire. It might be used to explore the magnetic field in a 

 rough way. I found, for instance, that when I brought up a 

 north pole on one side I could neutralise its effects by bringing 

 up a north pole to a proper distance on the other side. 



To get these sounds it is by no means necessary to have any 

 elaborate arrangement of stretched wire and so forth. If a 

 magnet be brought up to the w ire leading to the telephone, the 

 sound will be heard quite distinctly. If the wire be grasped 

 tightly in the fingers between the magnet and the telephone wire, 

 the sound is stopped, showing that it is transmitted mechanically 

 along the wire. This experimoit is certainly not new, but, 

 although I have seen the possibility of such action mentioned 

 (^■g. Wiedemann, " Galv.mismus," Bd. ii. p. 602), I have no- 

 where seen any indication that the sounds are so marked and 

 £0 easy to produce. I believe that this cause has been at work 

 along w'ith others in many experiments on the sounds obtained 

 in magnetisation ; for Instance, in De la Rive's experiments. 

 It is impossible, however, to decide with certainty, because 

 no sufficient indications are usually given as to the nature of the 

 magnetic field in which the wire conveying the interrupted current 

 was placed. 



The wire telephone arranged in this way with the wire in a 

 strong magnetic field is well suited for reproducing music. 

 Whether it could be adapted for articulate speech, I do not 

 know. 



The above experiments of course raise at once the question 

 whether the sounds in the ordinary wire telephone and those I 

 shall describe presently may not be due to the earth's mag- 

 netism. To settle this point, I stretched a brass wire 15 cm. 

 long in the telephone ; the wire was fine enough to give a 

 feeble sound of itself when the interrupted current of two 

 Bnnsen's cells was passed through it. I shifted the apparatus 

 about, so as to bring the wire as nearly as possible into the line 

 of dip, and then placed it perpendicular to that position ; but I 

 could not detect the slightest change in the intensity of the sound. 

 If it be borne in mind that here the distinction between wires 

 as to their thickness is only important in so far as it affects 

 their stiffness, it will, I think, be clear that this experiment settles 

 that the earth's magnetism is not an operative cause with the 

 current strengths I generally used. Another proof of this will 

 be given by and by. 



Effects due la the lilagnetism of the Telephone Wire itself. — The 

 following experiments were made with a view to test a conjecture 

 of Prof. Tait's, referred to in a letter to Nature, vol. xxii. 

 p. 168, and to settle, if possible, the cause of the exceptional 

 behaviour of iron wires in the experiments of De la Rive and Dr. 

 Ferguson. 



Two Bunsen's cells were used throughout, and the current was 

 interrupted by a tuning-fork driven by an auxiliary battery. 



My first experiment was made with an iron wire (a, 19 cm. 

 long, "50 mm. diameter). It gave a moderately loud sound to 

 begin with, a low note with a predominating fizz, not unlike the 

 fizz heard at the mercury cup in the far room (owing, I suppose, 

 to the volatilisation of the spirit by the heat of the spark, 

 which passes when the dipper of the tuning-fork leaves the 

 mercury). 



When a portion of the w ire was heated with a Bunsen flame 

 the sound increased very much for a short time, and then died 

 away again considerably after the wire got red hot. On allow- 

 ing the wire to cool, the sound, after a short time, suddenly 

 swelled out and then fell away again. The permanent sound 

 was, however, louder than it had been at first } 



I soon satisfied myself, by cautiou^ly bringing the flame up to 

 the wire, that there is a certain temperature at which the sound 

 is a maximum. The wire was heated up to white heat and 

 allowed to cool pretty rapidly, and it was found that the sound 

 was at its loudest at a dull red heat, jut before the phenomenon 

 of the re-glow occurred, along with which a peculiar crackling 

 could be heard, due, no doubt, to the abnormal contraction and 

 extension of the iron at that temperature. 



Several causes at once suggested themselves. The alteration 



* This phenomenon was observed by Dr. Ferguson independently, and 

 exhibited to the Royal Society of Edinburgh at the meeting before that at 

 which an abstract of the present paper was read. 



of the elasticity of the wire was dismissed as probably not the 

 principal cause at all events ; for the increase of the '^sound 

 begins at comparatively low temperatures. Although I did not 

 expect to find any such thing, I looked for a maximum of resist- 

 ance at a high temperature by placing the iron wire in one 

 circuit of a differential galvanometer, balancing it with an equal 

 resistance in the other, and then heating. I found, as is already 

 known, that the resistance increases with great rapidity after 

 dull red heat, but obtained no indication of a maximum. 



The most probable explanation seemed to be the magnetic 

 properties of the wire. It is well known that the magnetic 

 susceptibility of iron (that is, loosely speaking, its power to 

 become inductively magnetised under the influence of a given 

 magnetic force) is at its maximum about dull red heat ; that it 

 declines veiy rapidly at higher temperatures, and is almost in- 

 sensible at a bright red heat. The coercitive force of iron, that 

 is, its power to retain magnetism permanently, unaided by exter- 

 al magnetic forces, disappears at a much lower temperature. 



The sound in the above experiment depends, therefore, upon 

 temperature in the same way as the magnetic susceptibility of 

 the iron wire. This is strong proof that the sound is simply due 

 to the fact that the iron is magnetised. I convinced myself by 

 direct experiment that the effect extends throughout the whole of 

 the wire, for I found that two flames at different places produced, 

 when properly applied, 'more effect than one, and that, as I 

 brought more and more of the wire to the proper temperature, 

 the sound grew louder and louder. The fact that on cooling 

 the permanent effect was greater than before probably corre- 

 sponds to the fact that, under certain circumstances, the perma- 

 nent magnetism is increased by heating and subsequent cooling, 



Fig. 3. — Nickel. — A, buzz and low note, both loud : b, low note, gone ;_ c, 

 buzz, tending to soften ; D, buzz, gone ; e, feeble high note : f, high 

 note. 



of which fact it was easy enough to make certain by testing the 

 magnetism before and after heating. 



I next took a piece of steel piano wire (B, 20 cm. long, '9 mm. 

 diameter) and mounted it in the telephone. At first no sound 

 whatever could be heard. On magnetising it longitudinally, 

 by stroking once or twice ^vith a pair of _magnets, a sound 

 was heard quite distinctly, viz., a gentle fizz accompanied 

 by a high note. On magnetising more strongly this sound 

 became somewhat louder, but retained the same character. 

 Gentle heating wdth a spirit-lamp decreased the sound ; but it 

 recovered its intensity when the wire was allowed to cool, and 

 remained permanently a little louder than at first. Repeated 

 gentle heating and cooling increased the permanent sound 

 somewhat. 



The wire was then strongly heated with a Bunsen flame. At 

 first the sound died away to a minimum, then it increased, and 

 was very loud about a dull red, then it fell off again. When 

 the wire cooled the sound rose to a maximum, and then fell off, 

 no minimum being perceptible. After this the permanent sound 

 was a good deal louder, but the dimmution on slightly heating 

 could no longer be observed with certainty. 



I next heated the steel wne D to white heat throughout, so as 

 to deprive it entirely of all magnetism, and tempered it by 

 dropping it into cold water when dull red. When put into the 

 telephone after this treatment it gave no sound whatever. One 

 stroke with a pah: of bar magnets caused it to sound quite dis- 

 tinctly. It gave a gentle fizzing sound along v\ith a veiy high, 

 note. Repeated gentle heating and cooling gave the same 



