ELECTRIC WAVE TELEGRAPHY—FLEMING. A her a l 
the energy which can be stored up in it and liberated at each 
spark discharge is also small, at most a fraction of a foot-pound 
or a few foot-pounds. Hence it is rapidly frittered away by resist- 
ance and in radiation, and the oscillations are few, say half a 
dozen or so, and highly damped. If, however, we form an oscilla- 
tion circuit consisting of a large condenser, inductance and spark 
gap we can store up a larger amount of energy and liberate this sud- 
denly across the spark gap at each discharge. (See fig. 7.) If, then, 
these oscillations are made to induce others in a directly or induc- 
tively connected antenna, we can liberate the energy as radiation, 
and having a larger store to draw upon create longer trains, say of 20 
to 100 more feebly damped oscillations. 
Corresponding to these types of transmitter there are various suit- 
able forms of receiver. With a highly damped radiator we must use 
some form of wave-detector, such as a coherer, which is chiefly af- 
fected by the first or maximum oscillation, and this must be inserted 
in a receiving circuit which is easily set in oscillation by a single or 
at most a few electromag- 
netic impulses. On the 
other hand, this renders 
the receiver more liable to 
disturbance by vagrant an 
electric waves due to at- He 
= B 
mospheric electricity, or 
other transmitters if of 
sufficient strength. 
If, however, we employ 
a feebly damped radiator 
emitting long trains of 
waves, say 20 to 50 waves, 
we can make use of a 
stiffer receiver circuit, Fig. 7.—Antenna emitting feebly damped telegraphic 
that is one containing a toner 
good deal of inductance, and a detector such as Marconi’s magnetic 
detector, which operates under the action of feeble but oft-repeated 
and properly timed impulses. We have then the advantage that the 
receiving circuit can be made far less sensible to non-syntonic or 
isolated impulses unless these are of extreme violence. 
Again, there are certain forms of detector—such as the thermal 
and one of my own, to be described presently—which are affected by 
the product of the mean-square value of the oscillations during a 
train and by the number of trains per second. Hence, in this case 
the effect on such a receiver at a given distance under the same con- 
ditions will be increased by increasing the number of trains of oscil- 
Jations per second, as well as by diminishing damping in each train. 
Necwe es 
ey 
accom 
