On the succeeding nights I experienced similar disturbances, and it seemed as if they might possibly have been the prelude to several small shocks which occurred about this time (November 15, 16, and 17). On November 17, at 8 A.M., the needle was found in contact, and again at 5 P.M., and at 6 P.M. the shock of a small earthquake was felt _which caused a rattling sound in the telephone for about one minute after the motion had appeared to cease_. The needle swung considerably, but did not come in contact. The great objection to these observations is that it is possible that the movements and sounds which I have recorded might, with the exception of one case when the shaking was actually felt, possibly have been produced by causes other than that of the movement of the ground. To determine this I subsequently put up two distinct sets of apparatus to determine whether the motions of each were synchronous. So far as I went this appeared only to be sometimes the case:—but this is a question difficult to determine, unless a recorder of time be added to the apparatus. The greatest objection to observations of this sort is that the sensibility of the instrument is not constant. After a current has been running for several days it is no longer sensible to slight shocks, it appears as if its resistance had been increased. To overcome this it is necessary to resharpen the carbon points and bore out the pivot holes every three or four days. Farther, the battery varies. This might to some extent be overcome by using a battery with large plates. These two causes tend to reduce the sensitiveness of the galvanometer-like recorder—the deflection of the needle gradually becoming less and less, and therefore day by day needing a greater swing to bring it into contact with the iron. For reasons such as these this instrument, to be used successfully, appears to require considerable attention. Another form of microphone employed by the author consisted of an aluminium wire standing vertically on a metallic plate, its upper end passing loosely through a hole in an aluminium wire standard. The upper end of the vertical wire was loaded with lead. This contrivance possesses all the sensitiveness of an ordinary microphone, whilst, if it receives a sudden impulse, there is a sudden break in the current, and the vertical wire is thrown from one side to the other of the hole in the standard. After many months of tiresome observation with instruments of this description, and after eliminating all motions which might have been produced by accidental causes, the general result obtained showed that in Tokio there were movements of the soil to be detected every day, and sometimes many times per day, which to ordinary persons were passed by unnoticed.