MULTIPLEX TELEPHONY AND TELEGRAPH Y—SQUIER. 141 
D. C., this wire was connected directly to earth through a “perikon” 
crystal detector, such as is well known in wireless telegraphy, and a 
high resistance telephone receiver of about 8,000 ohms was shunted 
around the crystal. In this preliminary experintent no attempt was 
made at tuning, either at the transmitting end or at the receiving 
end of the line. 
In the primary circuit of the generator, arrangements were made 
by which either an interrupter and telegraph key or a telephone 
transmitter could be inserted by throwing a switch. 
In the line circuit a hot wire milliammeter was inserted in a con- 
venient position so that the effect of the operation of either the 
telegraph key or of the human voice upon the transmitter could be 
observed by watching the fluctuations of the needle of the milliam- 
meter. 
A loose coupling was employed between the two circuits at the 
transmitting end, and the line circuit adjusted by varying the coup- 
ling until the current in the line was 20 to 30 milliamperes. With 
this arrangement (1) telegraphic signals were sent and easily received, 
and (2) speech was transmitted and received successfully over this 
single wire with ground return. 
The ammeter showed marked fluctuations from the human voices 
and enabled the operator at the transmitting station to be certain 
that modified electric waves were being transmitted over the line. 
The actual ohmic resistance of the line apparently played an 
unimportant part for telegraphy at 100,000 cycles, since with one 
of the wires of the pair and a ground return, the effect of doubling 
the conductivity of the wire by joining both wires in parallel, although 
this arrangement increased the capacity of the wires, could not be 
detected with certainty by an operator listening to the signals and 
unaware of which arrangement was being used. 
Inserting in the line wire a noninductive carbon rod resistance 
of 750 ohms, which is practically the resistance of the line itself, 
could not be detected by any change in the intensity of the received 
signals. 
The next experiment was to determine what effect, if any, such 
sustained electrical oscillations would have upon the minute tele- 
phonic currents employed in battery telephony. 
DUPLEX TELEPHONY, USING ONE GROUNDED CIRCUIT. 
To determine the fact that electric waves of ultra sound frequency 
produce no perceptible effect when superimposed on the same circuit 
over which telephonic conversation is being transmitted, the next 
step was to use such a train of sustained oscillations as the vehicle 
for transmitting additional speech over the same circuit. For this 
