May 27, 1909] 



NA TURE 



SELF.CTIVE WIRELESS TELEGRAPHY. 

 A PAPER by Dr. Alex. Muirhead and myself, on some 

 ■^^ experiments and measurements in accurate wireless 

 tuning with open-circuit radiators, and the conditions under 

 which perfect selection is possible, was read to the Royal 

 Society in January of this year, and will appear in a forth- 

 coming issue of the Proceedings. 



The essence of it is that in signalling across land both 

 radiator and receiver must be completely insulated from, 

 and elevated above, the earth, if they are to be persistent 

 oscillators such as are capable of accurate tuning. Earth 

 connection damps out the vibration and spoils tuning : and 

 to get the best effect the lower capacity area must be not 

 only insulated, but must bo elevated above the earth until 

 its capacity with respect to the upper aerial is a minimum. 



To prove this, the received energy was measured at a 

 distant station by a Duddell hot-wire meter ; and several 

 series of measurements were taken with the lower capacity 

 at different heights above the earth, and also when con- 

 nected with the earth. 



The sensitiveness of a thoroughly tuned Lodge-Muirhead 

 system is extreme ; small power is sufficient, and the in- 

 ductive connection of the collector to the receiving instru- 

 ment may be separated by a surprising interval without 

 stopping communication. 



Under these non-earthed conditions every other station, 

 even near and powerful ones, can be tuned out and their 

 disturbance eliminated. 



Directly earth connection is made, tuning of the radiator 

 and collector is nearly gone, for they no longer have any 

 persistent free vibration period. Samples of a large 

 number of measurements are recorded in the paper. 



But from the paper as originally sent in an account of 

 the most striking experiment to illustrate the facility and 

 perfection of tuning on this system, when insulated capacity 

 areas are employed without any earth connection, was 

 accidentally omitted, though it has since been communi- 

 cated to the society. The experiment was made on May 

 14, 1907, and may be briefly described. 



Preliminary Information. 



Each aerial of the Lodge-Muirhead system consists of 

 a pair of capacity areas in the form of a couple of very 

 open " Maltese crosses " or squares of wire suspended 

 horizontally from .four posts like the framework of a 

 carpet, one above the other, and both well insulated from 

 the earth. Connection with each is made in the middle 

 by a special elaborately stranded cable to the instruments, 

 but no earth connection is made at all. 



A wheel coherer — revolving steel disc dipping into oiled 

 mercury — is employed as detector under the conditions of 

 accurate tuning ; or sometimes a point coherer, similarly 

 treated with oil. .An electrolytic coherer is even more 

 sensitive, but its leakage damps vibrations out and pre- 

 vents the accumulation ot impulses necessary for accurate 

 tuning, whereas the film of oil on the wheel coherer 

 insulates until the oscillations in the receiving tuned con- 

 denser circuit have mounted up sufficiently to break it 

 down and overflow through the detector. 



That is in brief summary the way signalling works, and 

 the following account has reference to signalling across 

 Kent between Elmer's End and Downe. 



Experiment in Diplex Telegraphy. 



.\t two stations, Downe and Elmer's End respectively, 

 the upper capacity area of each aerial w-as bisected 

 diagonally, the two triangular halves being insulated from 

 each other, and each connected to its own independent 

 receiving or sending arrangement. The lower aerial was 

 not bisected, but was doubled, an additional insulated 

 area being placed a few feet below the ordinarv one. By 

 this means each station was practically doubled, and the 

 two halves at each station made to correspond to a different 

 wave-length. 



Two senders at Elmer's End were then set to work 

 simultaneously, one to transmit the word " Liverpool " 

 continuously for a long time, the other the word 

 " steamships " continuously in the same way. Two in- 

 dependent receivers at the Downe station — one of them a 

 siphon recorder and one a telephone, though both might 



NO. 2065, VOL. 80I 



easily have been automatic recorders — each of them in- 

 ductively connected with one half of the aerial there, now 

 received simultaneously, one of them a succession of. 

 " Liverpools," the other a succession of " steamships, "^ 

 without the slightest confusion or interference or over- 

 lapping of any kind. 



In other words, diplex telegraphy (as distinct from, 

 duple.x) was found quite easy on this system of tuning, 

 which was specified by one of us in 1897. 



Experiment in Selection or Tuning Out. 



.\iiother experiment more recently tried is the following. 

 Two stations were arranged at Downe, 1200 feet apart,, 

 either of which could speak with great ease to Elmer 's- 

 End, and was strong enough to speak to a station thirty 

 miles away. One of the Downe stations was then switched: 

 on to " receiving," and both Elmer's End and the other 

 station at Downe were set speaking to it. 



The wave-length of one was 300 metres, of the other 

 660 metres, so as to compare Civil with Admiralty con- 

 ditions. 



By the mere motion of a handle the frequency of the 

 receiving station could be altered at will so as to corre- 

 spond either with the neighbouring 'sending station 1200 

 feet off, or with the distant sending station seven miles 

 off — which distance might, however, have been increased 

 immensely without any difficulty. A few trees intervenecf 

 between the neighbouring stations. 



In these circumstances, when properly adjusted, each 

 station could be heard separately ; that is to say, messages- 

 could be received first from one tuned-in station and then 

 from the other, without any disturbance from the station 

 tuned-out, although both stations were sending all the time- 

 strongly and simultaneously. The ease and large margin 

 with which selection could be achieved shows that the two 

 neighbouring stations could have been put still nearer, 

 while still retaining the power of complete tuning-out. 



Testing of Margin of Selection. 



Further experiments in the same direction were con- 

 ducted as follows : — - 



The two stations at Downe, 400 yards apart, were re- 

 arranged so that there were no trees between, only a few- 

 low hedges, thus making the test manifestly more severe. 

 A given power was then employed for sending at one of 

 these neighbouring stations, and the same power at the 

 distant Elmer's End station, while the other neighbouring 

 station was arranged for receiving from either of these 

 two at pleasure. Experiment was now directed to deter- 

 mine the conditions under which the neighbouring station 

 could be completely cut out, while still the distant one 

 could be clearly heard. In other words, to determine the 

 amount of separation between the primary and secondary 

 of the inductive connection which would eliminate all dis- 

 turbance from the neighbouring station adjusted to ordinary 

 commercial wave-length, while it would permit perfect 

 signals to be received on the siphon recorder from the 

 distant tuned station of longer or more nearly naval wave- 

 length. 



Case I. — Elmer's End sending with a wave-length of 

 580 metres. Neighbouring Downe sending with a wave- 

 length of 300 metres. The receiving Downe station was 

 attuned so as to cover a range of wave-length about 580- 

 metres on the average, but extending more than 20 metres 

 above and below. Under these conditions it was possible 

 completely to cut out the local station on a coupling of 

 35 inches, that is, with 3I inches separating primary and 

 secondary coil of the inductive connection ; w'hereas from 

 Eln-ier's End perfect signals could be obtained without dis- 

 'turbance on any coupling between 3j inches and 7 inches. 

 Indeed, as the exact pitch was reached at the receiving 

 adjustment, the signals received boomed out, as it were, 

 verv strongly. 



Case 2.— The Elmer's End wave-length was shortened" 

 to 510 metres, the local Downe station remaining at 300 

 nietres, and again a scries of readings was taken at the 

 receiving Downe station adjusted to 'an average of 510- 

 nietres wave-length. 



The coupling separation, which now just managed to- 

 cut out the local station, was 4 inches. Anything above 



