21G hertz's experiments. 



linear dimensions of the system, and for tlie most rapid vibrations two 

 spherical knobs, one charged positively and the other negatively, and 

 discharging directly from one to the other, have been used. 



Hertz in his original investigations used two plates about ^C'" square, 

 forming parts of the same plane, and separated l)y an interval of about 

 ()()""• Each plate was connected at the center of the edge next the other 

 l)late with a wire about 3(^"' long, and terminating in a small brass 

 knob. These knobs were within 1! or 3""" of one another, so that 

 when one plate was charged positively and the other negatively they 

 discharged to one another in a spark across this gap. An apparatus 

 about this size would produce waves 10 or 12 meters long, and its rate 

 of oscillation would be about 30,000,000 per second. As the vii)ration 

 actually produced by these oscillators seems to be very complex, the 

 rate of oscillation can only be described as '^ about " so and so. In a 

 subsequent investigation Hertz employed two elongated cylinders 

 about 15*='" long and about 3"" in diameter, terminated by knobs 

 about 4*^'" in diameter, and discharging directly into one another. 

 Such an oscillator produces waves from 00 to 70"" long, and conse- 

 quently vibrations at the rate of between 1(K),000,000 and .500,000,000 

 per second. Most other experimenters have used oscillators about the 

 same dimensions as Hertz's larger a})paratus, as the effects produced 

 are more energetic; but many experiments, es])ecially on refraction, 

 require a smaller wave to be dealt with, unless all the apparatus used 

 be on an enormous scale, such as could not be accommodated in any 

 ordinary laboratory. When we art^ thus aiming at rapid rates of vibra- 

 tion, it must be recollected that we can not at the same time expect 

 many vibrations after each impulse. If we havc^ a stiff spring- with 

 a small weight arranged so as to give a lot of its energy to the 

 surrounding medium, we can not expect to have very much energy to 

 deal with, nor many vibrations, and, as a matter of fact, we find that 

 this is the case. The total duration of a spark of even a large Leyden 

 jar is very small. Lord Rayleigh has recently illustrated this very 

 beautifully by his photographs of falling*drops ami breaking bubbles. 



We can not reasonably expect each spark to have more than from ten 

 to twenty effective oscillations, so that, even in the case of the slower 

 oscillator, the total duration of the spark is not above a millionth of a 

 second. It is very remarkable that the incandescent air (heated to in- 

 candescence by the sparlc) should cool as rapidly as it does, but there 

 is conclusive evidence that it remains incandescent after the spark 

 proper has ceased, and consequently lasts incandescent longer than the 

 millionth of a second. \V^hat is seen as the white core of the spark 

 may not last longer than the electric discharge itself, and certainly 

 does not do so in the case of the comparatively very slowly oscillating- 

 sparks thj)t have been analyzed into their component vibrations by 

 l)hotograi>hing them on amoving plate. The incandescent air remain- 

 ing in the path of such discharge is probably the conducting iiath 



