176 
FESSENDEN. 
enemy to read messages sent (since a code can never be re¬ 
lied upon absolutely for secrecy) would prove a source of 
trouble. 
Dr. Pupin has recently made a considerable improvement 
in the construction of lines for telegraphic and telephonic 
purposes. The formula for the rate of arrival of an electric 
signal in a cable was first given by Ohm, though it is com¬ 
monly attributed to Kelvin. Kirch off first showed that the 
velocity with which an electro-magnetic disturbance is prop¬ 
agated in a wire is equal to that of light, but this is an en¬ 
tirely different thing from the rate at which signals can be 
received, for this depends upon the time taken by the im¬ 
pulse to reach a readable value, i. e., upon the sensitiveness 
of the receiver and the electric constants of the cable. 
Various schemes have been proposed for increasing the 
speed of transmission. Heavyside showed, many years ago, 
that inductance could be used to oppose the effects of capacity 
in cables and give increased speed in signaling. Later, S. P. 
Thompson proposed to put this in practice. Reed suggested 
a modification of the method. The writer suggested and de¬ 
veloped the use of pure sine waves, a method which has 
been used by Crehore and Squire with considerable success. 
But the employment of inductances had never been thor¬ 
oughly worked out until Pupin took up the matter and 
in a masterly paper showed what conditions must be ful¬ 
filled to give good results. 
This invention will undoubtedly give increased speed of 
transmission in telegraphic cables and better speaking in 
telephone lines, and is of great value. The fact, however, 
appears to have been lost sight of that inductance can never 
thoroughly neutralize the effects of capacity, and that conse¬ 
quently it is at present premature to talk of telephoning from 
America to Europe. It has been overlooked that there are 
certain exponential terms in the equations of propagation 
through a single element of capacity and inductance, form¬ 
ing the suggested cable, and that, whilst the harmonic terms 
are not cumulative, the exponential terms are, and that the 
