The Early History of Seismometry (to 1900)
James Dewey and Perry Byerly
Early Studies of "Microseisms"
Virtually all of the early pendulum seismographs sometimes gave records of vibratory disturbances which lasted for hours and even days, and which seemed quite unrelated to earthquakes. We have alluded to the study of such small motions of pendulums in our paragraphs on Italian seismometry of the 1870's. It would be well to consider briefly the history of these investigations.
T. Bertelli (1873) has found studies of "spontaneous" pendulum motions as far back as 1643. He reports that the first investigators concluded that the motions were due to local or accidental causes, or were optical illusions. Bertelli began the first systematic observation of these motions in 1870 (De Rossi, 1877, p. 11). He noted a dependence of pendulum activity upon regional weather conditions, with greater activity occurring in winter and accompanying low-pressure zones. Bertelli believed that the pendulum motions were caused by tremors, from natural forces acting on a regional scale. This view was contested by a contemporary, P. Monte, who felt that the pendulum movements were due to local causes, such as air currents and cultural noise. De Rossi (1874b, 1875) has summarized the controversy between Bertelli and Monte. De Rossi undertook to explain the difference between the views of Bertelli and Monte, and he concluded that some of the motions were tectonic in origin (De Rossi, 1874a). He sensed a correlation between "tremor storms" and earthquakes. Largely at his urging, observatories were established in Italy to note the state of "endogenous activity", and the name "endogeneous meteorology" was applied to the study of the interior of the earth (De Rossi, 1883). The tremors at each station were observed with a "tromometer", consisting of a common pendulum whose motions were observed directly through a microscope. Daily maps were published showing tromometric activity throughout Italy. The project was abandoned, however, because of the difficulty of separating tremors due to tectonic forces from tremors due to air currents and traffic (Agamennone, 1906, p. 41).
Meanwhile, the "spontaneous pendulum movements" had attracted the interest of other scientists. The attempt of an English group to measure the lunar perturbation of gravity was thwarted because of the high level of background vibrations. Darwin (1882, 1883), who wrote the report of the group, has given a more complete history than we have been able to give of the early studies of spontaneous vibrations. Von Rebeur-Paschwitz (1894, 1895a) also carried out extensive studies of the motions exhibited by his horizontal pendulums.
John Milne began studies of "tremors" in 1879, in the hopes that earthquakes in Japan would be preceded by detectable noises, as the faults responsible for the earthquakes were preparing to break (Milne, 1882b). [The view that earthquakes originate from faults was referred to by Milne (1886b, p. 221) as the "ordinary supposition". De Rossi (1874a) had earlier hypothesized that waves were generated at the time of an earthquake by the lips of a "volcanic fracture" moving rapidly up and down with respect to each other. The Mino-Owari earthquake of October 28, 1891, with its spectacular faulting, helped convince Milne that faulting caused earthquakes by the release of strain energy which had been stored in rock through the slow deformation of the Earth's crust (Milne, 1898b, p 24-38). Another widely-held theory, suggested by Mallet (1862b) was that earthquakes were caused by water, rapidly vaporizing and condensing within the Earth. Milne used a microphone in his 1879 studies. His approach is similar to some used in today's renewed attack on earthquake prediction.] His views on the causes of "tremors" changed over two decades of study. In 1887, he suggested that the tremors were caused by wind and were propagated to areas where no wind was blowing (Milne, 1887b). In 1893, he altered his position, and held that tremors, or "earth pulsations", were caused by atmospheric pressure changes (Milne, 1893c). Milne did not specify whether the disturbances, once generated, propagated as elastic waves. He likened the pulsations to "the swells upon an ocean" (Milne, 1893c, p. 103-104). By 1898, however, Milne (1898b) had concluded that most spontaneous pendulum motions were caused by purely local phenomena, or were perhaps instrumental disturbances. The "tremors" studied by Milne had periods of several minutes. Milne gave convincing evidence that many were caused by small air currents around the instrument. Some were also thought to be due to the effect of climate on the soil outside the building in which the instrument was housed. For some disturbances, he had no sure explanation.
Figure 24. "Pulsatory oscillations" (period about six seconds) observed on September 6th-7th, 1898, on an Omori horizontal-pendulum seismograph in Tokyo (reproduced from Journal of the College of Science, Univ. of Tokyo, 11, plate IV).
Milne (1898b) was drawn to investigate the so-called "diurnal wave", a daily drift of the position of pendulum bobs which had also been studied intensively by von Rebeur-Paschwitz (1894, 1895a). Milne believed, with von Rebeur, that the drift was associated with the passage of the sun across the sky. He did not determine a definite cause for the diurnal wave, although he suggested that it might result from unequal evaporation of moisture from the ground surrounding the instrument.
One wonders how many of the pre-1900 investigators were actually observing the background elastic-wave noise which we today call "microseisms". Most of the early seismographs were very insensitive by today's standards, and to be recorded, micro-seismic motion would have had to be very large. Some of the most convincing seconds of "microseisms" were obtained by Omori (1899). The periods of these disturbances were between four and eight seconds, whereas the periods of the recording instruments were twice as long. The records reproduced by Omori (Figure 24) resemble the records of "storm microseisms" which we obtain at Berkeley today. But the amplitudes of the displacements in the waves were as great as 0.2 millimeters, an order of magnitude higher than the maximum microseismic amplitudes one would expect for such periods at a noisy site (Brune and Oliver, 1959).
From the Bulletin of the Seismological Society of America. Vol. 59, No. 1, pp. 183-227. February, 1969.