——EEE SS ——— eerlererrceeC r,OCUhre nnn ne eee 
A RETROSPECT OF WIRELESS COMMUNICATION. 553 
wiped out, so that the recoil would be dead beat, just returning to zero and staying 
there. That was the quickest possible method of discharge. But there would then 
be no oscillations. If the resistance were still greater than that, then the discharge 
would take longer to reach zero; it would degenerate into a leak, or at first into a 
sort of intermittent current, returning towards zero spasmodically in jerks all in the 
same direction. 
Now a flash of lightning is the discharge of a condenser, that is, the discharge of a 
store of electricity in the cloud ; and Dr. Simpson the eminent meteorologist has shown 
that the resistance to a flash of lightning usually exceeds the critical value which 
would make the discharge dead beat, and still more exceeds the value which would 
permit it to be oscillatory, and in fact makes it intermittent ; so that the lightning, 
if photographed, is seen to be a series of spits, succeeding one another very rapidly, 
and giving a jerky current allin one direction. Lord Kelvin’s theory of 1853 provided 
not only for the oscillatory discharge and its dead beat condition, but also for the leak 
likewise. If the resistance was enormous then the charge of the condenser would 
simply leak away, the law of discharge being then exactly like that of a hot body 
cooling. For heat has no inertia, and therefore has no tendency to make oscillations. 
The hotter the body the more rapid the leak, that is to say, the cooling process : and 
the return to zero is asymptotic, that is to say, the potential falls down an exponential 
or logarithmic curve, getting slower and slower as it comes nearer zero, and gradually 
approaching it, taking theoretically an infinite time actually to attain it. That 
theory of Lord Kelvin’s about electric oscillations or what he called transient currents 
dominated for me the nineteenth century, and was illustrated by innumerable experi- 
ments, at different times in the century. In the year 1889 I lectured on the Leyden 
jar at the Royal Institution, and showed many of these effects. 
But we did not know then that there was another reason for reducing the time 
during which the charge continued to oscillate. It was killed not only by the 
resistance of the circuit, but by a certain proportion of energy radiated away into 
space. We did not know that there was any such radiation, nor did Lord Kelvin. 
We knew, or might have thought, that such radiation was possible, by the analogy 
of a tuning-fork. A struck or excited tuning-fork gives sound vibrations which die 
out at a certain rate. They die out partly because of the resistance of the steel of 
the fork, and partly, indeed chiefly, if the fork is mounted on a sound-board, by 
reason of the radiation which is thrown out into the air. A genius might have surmised 
that, as the tuning-fork vibrates in air, so the discharge of a Leyden jar or other 
condenser, being a vibration in the ether, might possibly carve the ether into waves 
and emit energy in that way. That is what happens, but no one suspected it for a 
long time ; they did not know that the conditions for ether waves would be satisfied 
by an electric discharge. We had no sense for such waves, and could not tell that 
they were being emitted, even when we made the experiment. We were in the 
condition of a deaf person striking a tuning-fork or a bell. If yous could not hear 
the sound emitted by the fork you would not know that there was any; and you 
would certainly not experiment on the waves, measure their wave-length, and utilise 
them for purposes of communication. 
The first to show that an electric discharge would generate such waves, that is 
to say that an alternating or oscillating current would lose a certain fraction of energy 
to the ether at every swing, was George Francis FitzGerald, who in the year 1880 
examined the question mathematically, communicating it to Section A of the British 
Association ; and in 1883 followed it up with a further communication, in which he 
calculated the actual amount of energy lost in a second by a given condenser and self- 
induction. This was a great feat, and I will write FitzGerald’s result on the board, 
for it is used to this day. It showed that a short wave oscillator radiated much more 
vigorously than one that vibrated slower, that the radiation power, in fact, depended 
on the fourth power of the frequency, other things being equal. 
The radiating power of a current 7 oscillating in a time T in a circle of area 
Ta? is 
Sure! aN 
oe 5 (ra%i,) 
which simplifies to this in terms of wave length 
12(qra2i)2 
8x 10 alk Ci 
