Very similar devices to those shown are quite generally employed in all instruments with mechanical registration.
Two separate pendulums installed at right angles to each other are required to register both components of horizontal motion.
Photographic Much less massive horizontal pendulums can be employed to register photographically by the aid of light reflected from a mirror in the usual manner.
the needles were disturbed. The needles act like any other small, light pendulum and simply suffer certain mechanical agitation during the earthquake. In some less frequent cases there appear to be certain distinct magnetic effects that occur with the earthquake, but these are only imperfectly understood at the present time.
Interpretation of Serious com plications are likely to enter into the records of seismographs especially of the horizontal pen dulum. This instrument is more and more sensitive to minute tilting motions the greater its period of oscillation. If the ground at the Weather Bureau instrument tilts as little as only one inch in a mile the stylus of the seismograph will be thrown out of the line nearly one-quar ter of an inch. Furthermore, the instrument is sensitive to the twisting component of motion and finally we cannot feel entirely sure the steady mass has remained quite at rest. If the earthquake motion synchronizes with the free motion of the pendulum, the latter is quite certain to be set swinging. In the face of
these complications it is hardly possible, as yet, to interpret exactly the records obtained. Great improvements in the instruments are constantly being made and it appears as if the horizontal pendulums are likely to be displaced in the future by better types of apparatus.
The interesting features of the Weather Bureau record of the San Francisco earthquake appear in Fig. 3. The small pre liminary tremors began sat 8 h.— 19 m.— 20 s., A.M., that is 6 minutes 42 seconds after the strong motion recorded at the University of California. The stronger motion begins at "b*, on the record at Washington and at 8:32 to 8:35 the pen was deflected off the sheet by the very strong motion, producing a gap in the record which, however, was resumed when the motion subsided a little. The original record sheet is 36 inches long and vibratory motion is perceptible for fully four hours. The com plete period of the larger waves was from 20 to 30 seconds. Such vibrations are very slow and appear to be entirely imperceptible to human beings, nor do they affect buildings or other structures injuriously, so far as can be ascer tained. C. F. /immix.
Chief of the United States Weather Bureau.