R. S. HYER MEASUREMENTS OF ELECTRIC WAYES. 
67 
servation. Uniformity of conditions was insured by having an assistant, 
whose duty it was to determine whether the spark in a certain standard 
resonator would always he extinguished at the same distance from the 
vibrator. Unless this test showed regularity of action, no record was 
made of the distance at which the spark was extinguished in the resonator 
whose length was varied. Fig. 3 represents the results of five different 
sets of observations; the abscissae representing the circumference of the 
resonator, and ordinates the corresponding maximum distances to which 
a fixed spark could be carried before it was extinguished. The lower 
curve is for a long spark, and the upper one for a shorter one. By mak- 
ing the sparks very minute, they could be detected to a distance of 35 
meters, and possibly beyond. These extreme distances, however, were not 
convenient for the construction of these curves. The great difficulty en- 
countered in this work is the variable action of the vibrator; it may work 
well till the curve is half completed, and then suddenly fall in the in- 
tensity of its action. When the assistant with the standard resonator re- 
ports that he can no longer obtain sparks up to the former distance, the 
work of observation must cease, the blackened knobs of the vibrator must 
be polished and their distance adjusted so as to bring it back to its for- 
mer intensity. Sometimes it is found impossible to do this, and then the 
work of observation must all be gone over again. It will be observed in 
Fig. 3 that, while there are some differences in the several curves, they 
all agree in fixing the first maximum at about 220 c.m., the first minimum 
at 360, with a considerable value; and the second maximum at about 700, 
and considerably above the first; while a second minimum either lies at a 
considerable distance beyond 9 meters, or else has a greater value than 
the first. Reducing the various observations to a single line, the resulting 
curve is shown in Fig. 4 by line BY, while below it AX represents what 
was expected from my understanding of Hertz’*s views. About the only 
agreement between the two is the position of the first maximum. Why 
the first maximum lies at 360, and has such a high value, why the second 
maximum is greater than the first, why the second minimum does not 
correspond with the first, are all facts that I find it difficult to explain. 
That this form of curve is not peculiar to the particular vibrator em- 
ployed in this experiment is shown by the curves in Fig. 5, where the one 
already described is represented by the line BB, where AA represents that 
given by an oscillator with larger plates, and CO that obtained when 
small cylinders were substituted for the plates. In all of these the first 
minimum occurs sooner, and has a greater value, than we would expect 
from Hertz’s Anew. The cylinder, as was expected, showed less resonance 
effects than the plates, but all were decidedly more resonant than is 
deemed compatible with Thomson’s theory. The data for these resonance 
