EARTHQUAKES, Instruments for Re cording. The first record of an earthquake really worthy of being regarded as accurate and complete was obtained by Prof. James Ewing at the University of Tokio, on 3 Nov. 1880, by the aid of a new seismograph of his own in vention. This great epoch-making result was dimly foreshadowed, no doubt, by earlier crude attainments, yet Ewing's work was far in ad vance of that of the times, and his records were the first and only ones from which the approximate amplitudes, periods, and accelera tions of the motions of the ground could be deduced.
In order that a seismograph may faithfully record the motion of the ground during an earthquake it is in general a necessity that some portion of the instrument must remain relatively at rest throughout the disturbance. It seems to the writer that to Ewing more than any other is due the credit of having formu lated the kinetic basis of all modern seis mometry. He was the first to show how to realize and utilize the principle of the steady mass in the measurement of earthquake motion. There is scarcely any type of seismograph now in use that does not find its prototype among the large number of devices and arrangements either described or actually constructed and employed by Ewing at the University of Tokio from 1880 to 1883. We cannot attempt to de scribe in this brief account •any of his early work,' or the equally important contributions to seismology by Milne, Gray and others' of that period.
Within the past 25 or 30 years seismographs have multiplied very greatly throughout the civilized world, and by their records it is found that any considerable earthquake literally sets the entire crust of the globe vibrating and its tremors are measurable at almost any point on the surface by the aid of the sensitive Instru ments now employed.
Nature of Earthquake Motion.— A short statement explaining the motions of the ground will aid in understanding the instruments and their records. The commonly current concep tion that earthquake motion is a.comparatively• simple vibratory displacement of the ground in some particular direction .is shown to be de cidedly erroneous by the actual records. The motion, in fact, is exceedingly complex; the vibrations in a horizontal plane take place in all conceivable directions and are compounded with motion in the vertical. With this complex sort of vibratory displacements there may also coexist certain twisting and tilting rotations, No 'single instrument can possibly produce the record of such highly intricate motions. They must be separated or resolved-into elementary components for satisfactory registration. It is generally recognized that no less than six dis tinct and separate components require to be measured, namely, three displacements in direc tions at right angles to each other, that is, one north and south, one east and west and one vertical; also, three components of rotation about the same three axes. The turning on the
vertical axis may be called a twisting com ponent. The turnings on the horizontal axis are tilting effects. Seismologists have not yet succeeded in recording all of these elementary components. Unfortunately most of the instru ments thus far employed are influenced by more than one component of motion, and the records cannot therefore be entirely interpreted.
Seismographs.— The seismographs chiefly employed up to the present time are those de signed to register the horizontal vibratory dis placements of the ground. In a limited num ber of cases efforts have been made to register the vertical component of motion. In still fewer cases the tilting motions of the ground have been the subject of special measurements. The writer is not aware that the twisting component has thus far ever been recorded, although forms of apparatus for the purpose have been proposed.
Vertical Component.— Instruments for ver tical motion are not so easy to design, and the results obtained by their use have been less satisfactory than with other types either be cause the vertical motion of the ground is rela tively very small, or because it may be absent altogether. In seismographs for vertical motion the so-called steady mass must, in general, be suspended by means of very flexible springs, or their equivalents, 'in order that it may remain at rest in the vertical sense while the ground may rise and fall beneath it. If the springs are too stiff the steady mass fails to remain at rest but is made to rise and fall with the move ment of the ground. On the other hand, if the springs are too weak, very slight .variations of strength, due to temperature and prolonged ex cessive stress, cause the steady mass to wander slight amounts from its normal position of rest. As, the motions are generally greatly magnified, these slight progressive displacements often carry the tracing entirely off the sheet designed to receive it, or otherwise prove fatal to the de sired result. These characteristics of ordinary springs constitute serious limitations in the con struction of seismographs for vertical motion. What is -needed -in this lanuicutiOn is some sort of available material for springs that, under very great extension, does not exhibit with the lapse of time any minute slow, progressive, after deformations and is unaffected by variations of temperature.