DECEMBER 17, 1897. ] 
advantages. It maintains the surfaces of 
the spheres electrically clean, avoiding the 
frequent polishing required by Hertz’s ex- 
posed balls. It impresses on the waves ex- 
cited by these spheres a uniform and con- 
Fie. 2. Diagram of the Marconi Apparatus. 
stant form. It tends to reduce the wave 
lengths; Righi’s waves are measured in 
centimetres, while Hertz’s were measured 
in metres. For these reasons the distance 
at which effects are produced is increased. 
Mr. Marconi uses generally waves of about 
120 centimetres long. Two small spheres, 
a and 6, are fixed close to the large spheres, 
and connected each to one end of the 
secondary circuit of the ‘induction coil’ C, 
the primary circuit of which is excited by 
a battery E, thrown in and out of circuit 
by the Morse key K. Now, whenever the 
key K is depressed sparks pass between 1, 
2 and 3, and since the system A B contains 
a capacity and electric inertia, oscillations 
are set up in it of extreme rapidity. The 
line of propagation is Dd, and the fre- 
SCIENCE. 
893 
quency of oscillation is probably about 250 
millions per second. 
The distance at which effects are pro- 
duced with such rapid oscillations depends 
chiefly on the energy in the discharge that 
passes. A 6-inch spark coil has sufficed 
through 1, 2, 3, up to 4 miles, but for 
greater distances we have used a more pow- 
erful coil—one emitting sparks 20 inches 
long. It may also be pointed out that this 
distance increases with the diameter of the 
spheres A and B, and it is nearly doubled 
by making the spheres solid instead of hol- 
low. 
The Keceiver.—Marconi’s relay (Fig. 2) 
consists of a small glass tube four centi- 
meters long, into which two silver pole- 
pieces are tightly fitted, separated from 
each other by about half a millimeter—a 
thin space which is filled up by a mixture 
of fine nickel and silver filings, mixed with 
a trace of mercury. The tube is exhausted 
to a vacuum of 4 mm., and sealed. It 
forms part of a circuit containing a local 
cell and a sensitive telegraph relay. In its 
normal condition the metallic powder is 
virtually an insulator. The particles lie 
higgledy-piggledy, anyhow, in disorder. 
They lightly touch each other in an irregu- 
lar method, but when electric waves fall 
upon them they are ‘ polarized,’ order is in- 
stalled. They are marshalled in serried 
ranks ; they are subject to pressure—in fact, 
as Professor Oliver Lodge expresses it, they 
“ cohere ’—electrical contact ensues and a 
current passes. The resistance of such a 
space falls from infinity to about five ohms. 
The electric resistance of Marconi’s relay 
—that is, the resistance of the thin disc of 
loose powder—is practically infinite when 
it is in its normal or disordered condition. 
It is, then, in fact, an insulator. This re- 
sistance drops sometimes to five ohms, 
when the absorption of the electric waves 
by it is intense. It, therefore, becomes a 
conductor. It may be, as suggested by 
