But they are also doubtless due to the reverberation of very distant shocks. It is probable that there is not a minute of time without its earthquake somewhere, and the vibrations may often be transmitted to very great distances. In only a very few cases has it hitherto been possible to identify a tremor with a distant shock, and even then the identification is necessarily rather doubtful. One of the best authenticated of these cases was when M. Nyrèn, an astronomer at St. Petersburg, noticed on May 10 (April 28), 1877, a very abnormal agitation of the levels of his telescope, an hour and fourteen minutes after there had been a very severe shock at Iquique, in Peru.
Astronomers are much troubled by slight changes in the level of the piers of their instruments, and they meet this inconvenience by continually reading their levels and correcting their results accordingly. Of course, they also take average results. These troublesome changes are probably earth-tremors, with so slow a motion to and fro that the term tremor becomes inappropriate. This kind of change has been called a displacement of the vertical, since a plummet moves relatively to the ground. Thus, we found at Cambridge that as the pendulum danced it slowly drifted in one direction or the other. There was a fairly regular daily oscillation, but the pendulum would sometimes reverse its expected course for a few minutes, or for an hour. During the whole time that we were observing, the mean position of the pendulum for the day slowly shifted in one direction; but even after a voyage of six weeks the total change was still excessively small. How far this was a purely local effect and how far general we had no means of determining.
This is a subject which M. d'Abbadie, of the French Institute, has made especially his own. Notwithstanding his systematic observations, carried on during many years in an observatory near the Bay of Biscay, on the French side of the Spanish frontier, hardly anything has been made out as to the laws governing displacements of the vertical. He has, however, been able to show that there is a tendency for deflection of the vertical toward the sea at high tide, but this deflection is frequently masked by other simultaneous changes of unexplained origin.
This result, and the connection between barometric variations and earthquakes and tremors, should make us reflect on the forces which are brought into play by the rise and fall of the tide and of atmospheric pressure. Our very familiarity with these changes may easily blind us to the greatness of the forces which are so produced. The sea rests on the ground, and when the tide is high there is a greater weight to be supported than when it is low. A cubic foot of water weighs 62 pounds; thus if high-tide be only ten feet higher than low-tide, every square foot of the sea-bottom supports 620 pounds more at high than at low water; and 620 pounds to the square foot is nearly 8,000,000 tons to the square mile. Again, the barometer ranges
