372 



NA TURE 



[August i6, 1900 



receiver, and thus reproduce the original speech. It is 

 only necessary to connect the coils of the electromagnet 

 in series with a receiving telephone, and to cause the 

 steel wire on which the magnetic record has been made 

 to pass once again at the same speed and in the same 

 direction between the poles of the magnet ; for the 

 variation of the magnetic field which the wire produces 

 as it moves along will generate currents proportional to 

 the rate of variation of this field, and the telephone will 

 respond in the same way, and with the same degree of 

 accuracy, as did the receiving telephone in the early Bell 

 combination of a pair of magneto-telephones prior to the 

 employment of a rriicrophone and a battery. 



The diagrams in Figs, i and 2 show the arrangement 

 of the apparatus, c is the electromagnet on which are 

 the bobbins of wire B and b', which are connected either 

 to the transmitting or receiving instrument according as 

 it is desired to record a message or to listen to one 

 already recorded. Between the poles P and p' of the 

 electromagnet passes the steel wire F, which is wound in 

 a helix over two drums, T and t', to which the ends of it 

 are attached and which are driven by an electric motor. 



Fig. 2. 



SO that the wire winds off one drum on to the other. 

 The drums rotate at such a speed as to give the wire a 

 linear velocity between the poles of a metre per second. 

 At the same time the magnet C is moved at right angles 

 to the direction of motion of the wire, so that for one 

 revolution of the drums the magnet moves a distance 

 equal to the pitch of the helix. 



The message thus recorded can be effaced, according 

 to Herr Poulsen, by the simple process of passing the 

 steel wire between the poles of the magnet when the 

 latter is excited by a steady current from a battery. 

 This operation establishes in the wire a uniform magnetic 

 field at right angles to its length, and this field, so far 

 from interfering with recording a future message, appears 

 to be necessary before any such record can be made. 

 But if it be desired to keep the record, this may be done 

 instead, and the wire used over and over again to repeat 

 the same message. A thousand repetitions can be 

 made, it is said, without any diminution in loudness or 

 distinctness. 



If the recorded message is not sufficiently loud, it is 

 possible with a comparatively simple arrangement to 



NO. 1607, VOL. 62] 



greatly increase its loudness. For this purpose it is 

 necessary to arrange a series of parallel steel wires or 

 bands as shown in Fig. 3, all of which have been pre- 

 viously prepared for receiving a magnetic record by being 

 passed between the poles of a magnet excited by a 

 steady current. The wires are moved at a uniform rate. 

 The first wire, i, passes first between the poles of an 

 electromagnet M, which is connected to the telephone 

 line, and consequently receives a magnetic record of the 

 transmitted message. The wire next passes between the 

 poles of a second electromagnet. A, which is connected 

 in series with a similar magnet a', between the poles of 

 which the second wire, 2, passes. As the wire i passes 

 between the poles of the magnet A currents are induced 

 in the coils of this magnet, which, traversing also the 

 coils of the companion magnet a', produce a magnetic 

 record in the wire 2. A similar action occurs as the 

 first wire passes between the poles of the magnets B, 

 c, . . . z, which are connected in series with the 

 magnets b', c', z', so that there is established in each 

 of the wires 2, 3, 4, . . . «, similar magnetic records. The 

 wires 2, 3, 4, ... « pass finally between the poles of a 

 number of electromagnets, fi2> ^3' ^4) • • • O") which 

 are all joined in series with a telephone receiver, T. If 

 the two magnets which are joined together in series, 

 such as A, a', b, b', . . . are arranged in the same per- 

 pendicular to the direction of the steel wires, and if all 



— i 9 » ^ 



;- 2 



^o. 



..^Ll 



3' 



-^ 



Z' 



the wires are moved forward with the same velocity, then 

 the magnetisations in the various wires — 2, 3, 4, . . . // — 

 at points lying in the same perpendicular will be similar ; 

 and since all these points will pass at the same instant 

 between the poles of the magnets Q..^, Q3, O4, . . . fi», 

 they will there superpose their effects, and the intensity 

 of the sound in the telephone, T, will be increased in 

 proportion to the number of wires. It seems possible, 

 therefore, with this device to indefinitely increase the 

 loudness of the received message, and with quite a 

 simple arrangement to easily double or treble the 

 intensity of the sound. 



There is a second ingenious method by which it is 

 suggested that an increase in loudness can be obtained. 

 In this advantage is taken of the fact, already pointed 

 out, that the strength of the currents induced in the 

 coils of the receiving electromagnet depends on the rate 

 of variation of the magnetic field : it is therefore possible 

 to increase the strength of the currents induced by a 

 field altering from one given strength to another by 

 diminishing the time in which such alteration occurs. 

 If, therefore, we have a steel band on which a magnetic 

 record has been made, we may increase the strength of 

 the currents it will induce in the electromagnet connected 

 to the receiving telephone — and consequently the loud- 

 ness of the repeated message— by simply increasing the 



