APRIL 9, 1914] 
waves of all wave-lengths, and is inversely as the | 
square root of the product of the capacity and induct- 
ance per unit of length. Hence when a complex 
electromotive force, the result of speaking to a tele- 
phone transmitter, is applied to the end of a cable 
-the various simple harmonic waves into which they 
may be resolved travel along the cable with unequal 
speed and attenuation. The shorter waves travel 
fastest, but are worn out soonest. Hence the wave 
form is distorted by the disappearance of the higher 
harmonics and the resulting sound is enfeebled by the 
attenuation. 
Dr. Fleming proved these statements by a new and 
interesting experiment. A complex electromotive force 
comprising a fundamental wave having a frequency 
of about one hundred, and including higher harmonics 
of greater frequency was applied to one end of an 
artificial cable built on Dr. Muirhead’s plan, repre- 
senting a submarine cable fifty miles in length. By 
means of a Duddell oscillograph the wave form of 
this electric oscillation was projected on the screen. 
A second wire on the oscillograph was then employed 
to examine the current in the cable at various dis- 
tances, ten, twenty, thirty, etc., miles from the send- 
ing end, and to project on the screen a second curve 
representing the wave~form at various distances along 
the cable. It was seen that as the distance increases 
the wave form is reduced in height and smoothed out 
so as to show that the higher harmonics are gradually 
extinguished. In the case of a telephone cable this 
would mean that the received sound is not only fainter 
but altered in quality so that the syllable or word is 
no longer recognisable. 
Photographs were then shown, taken by Mr. Cohen 
at the General Post Office Research Laboratory, show- 
ing the distortion of various articulate sounds as trans- 
mitted through certain cables. A remedy for this 
distortion was first suggested by Mr. Oliver Heaviside, 
who proved mathematically more than twenty-five 
years ago that if the four constants of the cable were 
so related that the quotient of the inductance by the 
resistance was equal to the quotient of the capacity 
by the leakance, then waves of all wave-lengths would 
travel at the same speed and attenuate at the same 
rate. 
In all ordinary cables the first-named quotient is 
much smaller than the second. Hence to remove 
distortion we may either increase inductance or leak- 
ance. Heaviside suggested increasing the former, 
and Prof, Silvanus Thompson in 1891 suggested in- 
creasing the latter by providing the cable with in- 
ductive leaks. Practical telephone engineers preferred, 
however, to decrease the resistance of the cable by 
increasing the copper section so far as possible. There 
is, however, a limit to this from the point of view 
of cost. Also the invention of paper-insulated cables 
for telephony assisted matters by reducing the capacity 
of the cable. Nevertheless a very important advance 
was made by Prof. Pupin, of Columbia College, New 
York, in 1899 and 1900, when he proved that Heavi- 
side’s suggestion could be put into practical form by 
loading the cable with coils of wire wound on iron 
wire cores inserted at equal intervals, but so close that 
at least eight or nine coils are included in the distance 
of one wave-length of the average wave frequency 
which is always taken at 800. If the coils are placed 
farther apart relatively to the wave-length they do 
more harm than good. Dr. Fleming illustrated this 
by a very pretty experiment of his own consisting of 
a string loaded at intervals with beads, one end of 
the string being fixed and the other twirled round by 
a motor so as to produce on it stationary waves. 
When the half wave-length was adjusted to be nearly 
equal to the distance between the beads, the cord 
refused to transmit the oscillations. 
Neteearo:: VOL 7826: 
NATURE 
\ 
151 
It was also illustrated by the production of stationary 
electric waves on a series of helices of wire having 
loading coils, or coils of high inductance introduced 
at intervals. 
An experiment was also shown with an artificial 
cable representing forty miles of standard cable into 
sections of which loading coils could be introduced or 
cut out as required. It was shown that when the 
cable was loaded the current flowing out of it at the 
receiving end was greatly increased when constant 
electromotive force was applied at the sending end. 
It is found then that loading telephone wires by 
suitable coils of high inductance placed at proper 
intervals of a mile up to ten or twelve miles accord- 
ing to the cable, greatly reduces the attenuation of 
the waves, although it is difficult to add sufficient 
inductance to cure distortion completely. 
Dr. Fleming gave a mechanical illustration of this 
effect. He said, suppose two similar ships were to be 
launched together side by side down ways of equal 
inclination and allowed to glide out into the sea as 
far as they would go until brought to rest by friction 
of the water. If then one of the ships was loaded 
with ballast so as to make it much heavier than the 
other, then, although entering the water with the 
same speed, the heavily loaded ship would glide out 
further than the other because it would possess a 
greater store of kinetic energy. So it is, he explained, 
with the electric waves on wire. By adding induct- 
ance to the circuit the wave energy is increased, and 
the waves attentuate less for a given distance of 
travel. 
This proposal of Pupin has proved to be a very 
practical solution of the problem of reducing the 
attenuation of telephonic waves. Both aerial lines, 
underground cables,. and submarine cables can be 
‘loaded’? or ‘‘Pupinised’”’ by inserting appropriately 
made inductance coils at equal distances, and the 
result is to reduce the attenuation to half or less than 
a half of that of the unloaded cable, and therefore 
to reduce the enfeeblement of the sound. 
In the case of aerial lines there is no difficulty in 
inserting these loading coils in the run of the cable. 
The coils are contained in iron boxes attached to the 
telegraph posts at intervals of six to twelve miles. 
The coils themselves consist of an iron wire core 
wound over with wire, and have generally an induct- 
ance of about o-2 henry, and a resistance of 6 or 
8 ohms. In the case of underground cables the load- 
ing coils are placed in pits at intervals of two or three 
miles. Such underground cables consist now of 
paper-insulated double metallic circuits; a large 
number of such circuits being included in one water- 
tight lead sheath. The problem of loading a sub- 
marine cable was more difficult to solve because the 
insertion of heavy iron-cased coils was out of the 
question. The cable had to be loaded in such manner 
as not to thicken it up inordinately at any point, and 
to permit of its being laid in the usual manner and 
lifted again if necessary for repairs. This particular 
problem was solved by Messrs. Siemens Bros. by the 
invention of a particular form of cylindrical loading 
coil which could be inserted in the run of a cable of 
the usual double-circuit type at distances of one 
nautical mile or so. When once it had been shown 
that such loading was effectual, telephonic engineers 
in all countries began to adopt it. In the United 
States the American Telephone and Telegraph Com- 
pany has equipped with loading-coils lines up_ to 
2000 miles in length. The longest aerial loaded line 
is that from New York to Denver. It is composed 
of No. 8 hard-drawn copper wires, the circuits being 
twisted to avoid cross talk and loaded every eight 
miles with coils having an inductance of 0-265 henry 
(see Table I.). 
