|Share page||Visit Us On FB|
34 HOPI SONGS
ment. At so slow a rate as 135 to 140 revolutions per minute, the motor could no longer perhaps be depended upon to give the cylinder a constant speed. The trial of different rapidities confirmed this conclusion. The electrical supply being reduced until the cylinder made on starting but 72 revolutions a minute, the note wandered over a wide interval, during most of the reproduction a minor third, but at the end becoming a sixth. At 120 revolutions the variation, while unceasing, kept within comparatively narrow limits, perhaps a quarter tone or somewhat more. In trying the same cylinder again on another day at 135 revolutions, I estimated the variation at nearly if not quite a semitone from first to last. Sometimes a change up and back through perhaps an eighth of a tone took place in a few seconds, but the general rule seemed to be a change to a new level, which it thereupon held for some time. At 158 revolutions the same cylinder showed a total variation which I estimated at less than a quarter tone, and perhaps not much more than an eighth,—certainly within an eighth during much the larger part of the cylinder. After trying 120 revolutions on the day before mentioned, I set the instrument at 180, the result being that the note seemed to the ear quite steady.
Since at this speed the pitch of the record on Cylinder A was far removed from Ut3, I took on the day after an inscription of the harmonium c', at 176 revolutions, Cylinder B. In order to make the reproduction beat with Ut3 with sufficient slowness to be easily counted, this rate had to be reduced to 170. At 174 revolutions I found the needle covered the cylinder to the very end in just two and a half minutes. We may then take 165 to 175 revolutions as the standard speed, and Cylinder B as a standard cylinder both in the inscription and when it was reproduced to beat with Ut3.
The beats became most distinct when the auxiliary tube was introduced into the resonator of the fork, a minimal tone being produced from the latter by gentle taps with the finger near the junction of the tines. Under these conditions I examined a quadruple series of reproductions from B, the regulator screw being untouched meanwhile, and the time between one and the next being only that required to replace the diaphragm at the head of the cylinder. As the needle took something over two minutes to traverse the inscription, this was a nearly continuous test of the performance of the instrument during a lapse of about ten minutes. The number of beats in those ten second spaces in each reproduction during which I was able to count them were as follows: —