ELECTRICAL PRINCIPLES OE THE ATLANTIC TELEGRAPH. 425 
the cable is alternately connected with the battery and the 
earth at regular intervals, its electral condition is changed at 
each contact to an extent corresponding to the electromotive 
force of the battery ; but if the contacts succeed each other at 
more than a certain rate, no alterations whatever are per- 
ceived at the other end of the wire, but a current appears 
there of constant strength, equal to that which would be 
produced by a battery of half the power if kept continuously 
connected with the cable. 
It is evident, therefore, that, whatever may be the delicacy 
of the receiving instrument, there is for each cable an absolute 
limit to the number of legible signals which can be thus 
transmitted in a given time. If this limit be exceeded, no 
matter how great the battery-power employed at the sending 
end may be, and the consequent extent of the changes pro- 
duced in the electrical condition of that end of the wire, no 
variations whatever are produced in the strength of the cur- 
rent at the receiving end. Within this limit, however, the 
effect of alternate battery- and earth-contacts at the sending 
end is to cause oscillations in the strength of the current at 
the receiving end. By making the alternate contacts more 
and more slowly, the extent — or, as it would be called in more 
scientific language, the “ amplitude” — of these oscillations 
can be increased, until the received current varies from one 
of imperceptible strength to the strongest the battery is 
capable of sending. From this it will be seen that the 
number of signals transmitted cannot be increased beyond a 
certain point, except at the expense of their legibility ; and, 
as we have already stated, there is another limit of speed 
beyond which they become altogether obliterated. 
The precise conditions which determine the speed with which 
distinct signals can be made to follow each other through a 
submarine cable were pointed out, by Professor William Thom- 
son,* eleven years ago. Professor Thomson showed that the 
intervals which must be allowed to elapse between successive 
signals, in order that they may have any given proportion of their 
greatest possible distinctness, is proportional to the electrical 
resistance of the conductor, to its electrical capacity, and to the 
square of its length. Hence, if cables of the same make were 
laid between Dover and Calais, and between Valentia and 
Newfoundland, their relative length being as 26 to 2,100, or 
as 1 to 80 nearly, the greatest rate at which it would be pos- 
sible to signal through the latter cable would be about 6,400 
times slower than the greatest rate possible with the former. 
Hence, also, if the length of any proposed telegraphic cable 
* Proceedings of the Boyal Society , vol. vii., 382 (May 24th, 1855). 
