120 
WIRELESS TELEGRAPHY. 
the base of the building, and connected with either side of that gap it 
served to bring into action the energy developed by the dynamo, and 
everybody saw the electric arc which it produced. 
The wave can be detected by a minute spark only visible by 
careful observation, or by a vacuum tube ; but by far the most sensitive 
arrangement is the " coherer” to which I now draw your attention. The- 
spark arrangement would perhaps act at about double the distance I 
have it (about 4 feet) but not further; the distance at which the 
vacuum tube will perceptibly glow is smaller still; but the coherer is ex¬ 
tremely sensitive to these electro-magnetic waves. It consists mainly 
of a glass tube in which are two silver plugs, separated by a very short 
distance, the space between them being half filled with metallic filings, 
this little tube of glass with the metallic filings is the soul of the Marconi 
apparatus. 
(The coherer shewn ; and a few distributed among the audience). 
In order to explain its action I must make a slight digression. 
Some time ago the discovery was made that a tube filled with metallic 
filings in a finely divided state offers a large resistance to the electric 
current—in fact it is practically a non-conductor. Possibly each par¬ 
ticle is sorrounded by a film of gas. 
Then experimenters after Hertz noticed the curious fact that 
a tube such as I have described immediately becomes conducting in a 
greater or less degree when Hertzian waves impinge on it, the drop 
in resistance depending on the intensity of the radiation falling upon 
the tube. It was found that this form of detector was extraordinarily 
sensitive. The simplest way of using such an instrument to detect 
waves would be to connect a single voltaic cell with a galvanometer 
in circuit with it. Owing to the high resistance of the coherer 
little or no deflection is produced on the galvanometer until a 
wave of sufficient intensity impinges on the coherer. Immedi¬ 
ately this occurs the resistance of the coherer drops, and a current) 
flows showing a deflection on the galvanometer. 
(Experiment shewn :—Coherer suspended across the gap of the re¬ 
ceiving instrument and placed in simple circuit with a single voltaic 
cell and galvanometer. Immediately a spark took place across the 
transmitter, the galvanometer showed a full deflection). 
It will be observed that the “ cohersion ” remains after the exciting 
cause is removed, and the galvanometer will shew a deflection for an in¬ 
definite time, unless the original non-conducting state of the coherer 
is restored. This can be done by mechanically shaking up the filings 
with one or two smart taps. The receiver is then in a position to re¬ 
cord a second signal. 
(Experiment shewn : The coherer was tapped and the galvanometer 
needle again returned to zero). 
You see that the coherer is the soul of the apparatus. Now 
the problem was to make use of the coherer to give audible signals, 
and in order to do that, Mr. Marconi had to arrange an automatic 
method which, when the signal is over, caused the coherer to be 
tapped and the signal to cease. What happens in the coherer is a 
