distance might be raised to half a mile.

As early as 1879 Professor D. E. Hughes began a series of experiments in
wireless telegraphy, on much the lines which in other hands have now
reached commercial as well as scientific success. Professor Hughes was
the inventor of the microphone, and that instrument, he declared,
affords an unrivalled means of receiving wireless messages, since it
requires no tapping to restore its non-conductivity. In his researches
this investigator was convinced that his signals were propagated, not by
electro-magnetic induction, but by aerial electric waves spreading out
from an electric spark. Early in 1880 he showed his apparatus to
Professor Stokes, who observed its operation carefully. His dictum was
that he saw nothing which could not be explained by known
electro-magnetic effects. This erroneous judgment so discouraged
Professor Hughes that he desisted from following up his experiments, and
thus, in all probability, the birth of the wireless telegraph was for
several years delayed.[3]

[Illustration: Fig. 71.--Marconi coherer, enlarged view]

The coherer, as improved by Marconi, is a glass tube about one and
one-half inches long and about one-twelfth of an inch in internal
diameter. The electrodes are inserted in this tube so as almost to
touch; between them is about one-thirtieth of an inch filled with a
pinch of the responsive mixture which forms the pivot of the whole
contrivance. This mixture is 90 per cent. nickel filings, 10 per cent.
hard silver filings, and a mere trace of mercury; the tube is exhausted
of air to within one ten-thousandth part (Fig. 71). How does this trifle
of metallic dust manage loudly to utter its signals through a
telegraphic sounder, or forcibly indent them upon a moving strip of
paper? Not directly, but indirectly, as the very last refinement of
initiation. Let us imagine an ordinary telegraphic battery strong enough
loudly to tick out a message. Be it ever so strong it remains silent
until its circuit is completed, and for that completion the merest touch
suffices. Now the thread of dust in the coherer forms part of such a
telegraphic circuit: as loose dust it is an effectual bar and obstacle,
under the influence of electric waves from afar it changes instantly to
a coherent metallic link which at once completes the circuit and
delivers the message.

An electric impulse, almost too attenuated for computation, is here able
to effect such a change in