May 12, 1910J 



NATURE 



0^0 



y the action of the local current passing through the 

 otitact and round the winding K. It is by the action of 

 tie local current operating through this winding that the 

 onduction space is formed and afterwards maintained, 

 k) good is the automatic adjustment that the instrument 

 :iay be turned upside down, producing hardly any notice- 

 ble alteration in the value of the local current and with- 

 ut any effect on the working of the relay. The regu- 

 ating winding K must not act when traversed by the 

 apidly varying telephonic currents ; this is brought about 

 »y surrounding the iron under the coil by a closed circuited 



I'/aV 





Fig. 3. 



copper sheathing. Eddy currents set up in this sheathing 

 by mutual induction destroy the self-induction of the coil. 



Fig. 2 is an enlarged view of the reed P and the contact- 

 pieces M, O. In the present instrument the contact is 

 made between metal pieces of hard osmium iridium alloy. 

 The top contact is pointed ; the lower one is flat, and is 

 soldered to the reed ; both are polished, and work under a 

 small drop of thin oil. 



In earlier instruments the lower contact O was carried 

 by a thin iron disc ; the relay was then very susceptible to 

 outside noises. For this reason a reed is now used ; it 

 exposes such a small surface to the air that it is practically 

 unaffected by extraneous sound. 



Fig. 3 shows the connections of the relay. C is a Ary 

 cell (this is the normal voltage, which is as high as it is 



currents. Speech or signals that are too faint to be heard 

 in the ordinary Bell receiver may be heard clearly through 

 the relay. If a watch be held against the ear-piece of a 

 Bell telephone the induced currents produced when passed 

 through the instrument will reproduce the ticking in the 

 receiver attached ; this is a severe test. 



This property of magnifying feeble telephone currents 

 has made it particularly useful in wireless telegraphy. On 

 replacing the telephone by the relay the increased sensitive. 



Yxc. 4. 



desirable to employ). K the low resistance regulating wind- 

 ing, T the receiving telephone or telephone head-piece of 

 approximately 40 ohms resistance, D is an amperemeter 

 or current indicator ; when the microphone contact is 

 opened so as to cut down the local current to half its 

 maximum value, the relay is usually at its best adjust- 

 ment. The telephone currents to be magnified enter bv 

 the terminals marked A, and circulate through the wind- 

 ing H. 



The relay will magnify the very feeblest telephone 



NO. 21 15, VOL. 83I 



Fig. 5. — The Telephone Re'ay. 



ness thus obtained doubles the distance over which it is 

 possible to receive signals. Its utility in this direction has 

 been tested, among others, by the -Admiralty and the Post 

 Office. In a wireless receiving station, messages, the verj' 

 existence of which was not even suspected, owing to their 

 extreme feebleness, when listened for under former con- 

 ' ditions with the relay in circuit, were easily read. At 

 the invitation of Mr. Marconi I took two instruments to 

 the Haven Hotel, Poole. In one of the tests 

 (Clifdon, in Ireland, sending with the Marconi 

 musical spark) the signals were heard in the 

 telephone, directly connected, as a faint but 

 clear and pleasing series of musical notes ; but 

 with two relays joined to the system and work- 

 ing in series the notes were rendered so loudly 

 as to be heard clearly by everyone in the room, 

 and an operator listening at a distance of 

 several yards from the instrument could have 

 deciphered the message. The relay is not 

 easily affected by extraneous noises and vibra- 

 tion. It can thus be carried on board ship 

 and worked in all weathers. 



.As regards its utility on ordinary telephone 

 lines, speech may be magnified many times in 

 loudness without perceptible loss in the articu- 

 lation, and it will work with large currents 

 to a point at which the Bell receiver in its 

 local circuit is responding with uncomfort- 

 able loudness. In experimenting over a 

 20-lb. standard cable and speaking only 

 one way. it has been proved that, when the relay is 

 applied, thirty miles may be added to any length through 

 which it is possible now to speak direct. For instance, 

 supposing the length of the core for direct speaking be 

 twenty miles, this may be increased to fifty miles for the 

 same loudness and approximate clearness when the relay is 

 in circuit, either as a single repeater at the end of the 

 first twenty miles or as a receiver at the end of the .fifty 

 miles. 



These tests prove that the telephone currents must be 



