The Early History of Seismometry (to 1900)

James Dewey and Perry Byerly

The Invention of a Seismograph in Italy

The research begun by Cavalleri and Palmieri was vigorously continued by Italian seismologists in the 1870's. 1874 saw the publication of the first journal devoted to solid-earth geophysics, the Bulletino del Vulcanismo Italiano, founded and edited by M. S. De Rossi. In spite of the high seismological activity, however, relatively few instrumental advances emerged from this period, and the instrument which to us is most significant seems to have been little noticed at the time it was invented. This was Cecchi's seismograph, which we will describe in the next paragraphs. The great majority of the other instruments designed by Italian seismologists at this time were seismoscopes of the same general type as had been invented earlier by Palmieri.

Cecchi's instrument has been described by later Italian seismologists as the first true seismograph (Agamennone, 1906, p. 91). The machine was apparently built in 1875 (Cecchi, 1876). Unlike any of the instruments we have discussed so far, the Cecchi seismograph was expected to record the relative motion of a pendulum and the Earth as a function of time. For horizontal vibrations, two common pendulums were used, one vibrating in a north-south plane and the other vibrating in an east-west plane. The pendulums "beat seconds"; their motion was magnified three times by a thread-and-pulley apparatus. For vertical motions, a mass on a spiral spring was used. Finally, a machine which was expected to record rotary motions, was incorporated into the seismograph. This consisted of a cross bar with weights at both ends, much like a dumbbell, which was pivoted at its center of mass so as to rotate in a horizontal plane. Restoring force was applied to the dumbbell by springs, so that it oscillated with a period of one second.

Cecchi arranged a seismoscope to start a clock and to start into motion the recording surface at the time of an earthquake. The recording surface would translate under the indicating needles at a speed of one centimeter-per-second for twenty seconds. From the time on the clock, an observer arriving at the seismograph would determine how long before his arrival the earthquake had occurred.

Unfortunately, the subsequent history of Cecchi's earliest seismograph is unclear. A modified form of the instrument was reported to have been installed in Manila (Du Bois, 1885). This instrument and the later Cecchi seismographs reported by Agamennone (1906) are not so interesting to us because, by the time these instruments were built, better seismographs were being used by British scientists in Japan. The early form of the Cecchi instrument was apparently installed in several observatories. One might think that one of these seismographs would have recorded an earthquake in the five years between 1875 and 1880, the latter being the year of the earliest seismogram obtained by the British in Japan, who also claimed the first real seismograph. But the earliest date we have found for a seismogram obtained with the early Cecchi instrument is February 23, 1887, when one of these seismographs recorded a large earthquake which occurred in the French-Italian border region (Denza, 1887). In that earthquake, only the east-west component was recorded, although the seismograph was located in a zone of strong shaking. The instrument must have been most insensitive. The seismogram is reproduced in Figure 6. The record appears to us to be as accurate a representation of earth movement as was obtained by the early seismographs in Japan. Assuming that the recording instrument had not been significantly altered between 1875 and 1887, it would seem that Cecchi indeed deserves credit for the construction of the earliest seismograph, insensitive though it might have been.

Figure 6. The record obtained by a Cecchi seismograph at Moncalieri, Italy, on February 23, 1887 (reproduced from Fouqué, Tremblements de Terre, Bailliére, p. 79).

Cecchi's seismograph had relatively little impact upon Italian seismology. The insensitivity of the instrument must have discouraged others from building seismographic devices (De Rossi, 1887). In addition, Cecchi's apparatus would have been costly and unsuitable for an observatory of modest means. Finally, many Italian seismologists seemed to believe that an earthquake could be satisfactorily described by seismoscopic data (De Rossi, 1877, p. 9). We won't discuss in detail any more of the Italian seismoscopes. Descriptions of many are given by Agamennone (1906) and Ehlert (1897a). A review in English of the Italian instruments was given in The Electrical World (anon., 1887). Some of these seismoscopes depended on the fall of a delicately-balanced object to trip a time-recording device and sound an alarm. Many were modifications of Palmieri's seismoscopes. The microphone was used by some Italian seismologists to listen to "earth noises" (De Rossi, 1883).

An example of the use of seismoscopes in Europe is the experiment of von Lasaulx (Hoernes, 1893). He constructed an apparatus in which the fall of a poised weight caused a pendulum clock to be stopped. About one hundred and fifty of these devices were installed at telegraph stations in Germany. The sensitivity of the installed seismoscopes was deliberately kept low, so that human disturbances would not cause the clocks to be stopped frequently. Nevertheless, two earthquakes were large enough to be detected by many of the seismoscopes. Time and "direction of motion" were determined from each instrument. The data were contradictory and seemed only to indicate that most of the clocks at the telegraph stations were not sufficiently accurate for seismological work.

From the Bulletin of the Seismological Society of America. Vol. 59, No. 1, pp. 183-227. February, 1969.