ON SEISMOLOGICAL INVESTIGATION. 273 



measure tilts or displacements in the plane of the prime vertical : the 

 mirror being in this plane, that of the suspending wires being at right 

 angles. Disturbances are shown by the motion of a spot of light being 

 carried to the east or west of its normal position. The more noticeable 

 deviations from uniformity have been reported to Professor J. Milne as 

 produced probably by earthquake shocks ; the dates of these interruptions 

 appear in another part of the report. The instrument, however, does not 

 seem very well adapted for the measurement and discussion of these irre- 

 gular motions. Some difficulty arises from the smallness of the time scale 

 (10 m.m. of paper passing in an hour), in consequence of Avhich small rapid 

 vibrations are indistinguishable, while the sensitiveness of the instrument 

 as at present used does not seem to be sufficiently great to record the 

 characteristic motion. In Dr. Davison's instrument a displacement of the 

 spot of light through 3-44 inches corresponded to a tilt of the mirror of one 

 second of arc. I endeavoured at first to reach this degree of sensitiveness, 

 and met with difficulties, some of which will be mentioned later. But a 

 recommendation was sent to me last May by Professor Milne to so arrange 

 the instrument that a tilt of the ground of two seconds should move the 

 spot of light on the scale 1'74 inch. I have endeavoured to conform to 

 this direction with the result that the instrument appears more stable, 

 and the trace produced on the sensitised paper seems still more adapted to 

 the discussion of the bending of the earth's crust throughout considerable 

 periods of time than for the observation of the pulsations produced by 

 irregular and violent shocks. The following discussion is therefore 

 confined entirely to the uniform behaviour of the spot of light. 



For a very considerable time after the instrument was mounted the 

 photographic trace showed such a continual and rapid motion towards 

 the east that all other effects were completely masked. This was no 

 doubt due to a want of stability in the steel rod carrying the instrument, 

 which had not yet come to a position of rest. The frequent alteration of 

 the foot-screws, always in one direction, necessary to bring the spot 

 of light back on the scale, disturbed the level to such an extent that the 

 sensitiveness varied considerably. This was no doubt assisted by a motion 

 of the rod in the plane of the meridian, which would alter the horizontal 

 distance between the points of suspension of the wire carrying the mirror : 

 a motion which would not be visible on the photographed trace. About 

 November the constant motion of the mirror became less recognisable, and 

 the series of measures here described was begun in January of this year, 

 at which time it is hoped the instrument had settled into its normal con- 

 ditions. The scale for converting the linear displacement of the photo- 

 graphic trace into seconds of arc still continues, however, to give some 

 trouble, and this feared irregularity in the scale has prevented the use of 

 many of the records in the following discussion. The scale, it may be as 

 well to explain, is determined by turning the mirror through a known 

 angle by means of a rocking-arm, capable of being moved from a distance 

 by the alternate inflation of one or other of two india-rubber balls. It 

 lias been the rule to move this rocking-arm once a day, and those records 

 are considered trustworthy when the linear displacement of the light is 

 the same at the beginning and end of the day. When the displacement 

 is not accordant there seems to be no way of making the observations 

 available for discussion. Moreover, it is found practically that very 

 different intervals of time are required to bring the mirror to a state of 



1898. T 



