ON TUB EAKTHQTTAK.E PHENOMENA OF JAPAN. .211 



which vibrations of various descriptions were propagated. The arrange- 

 ments were automatic. The charges were fired electrically, and as the 

 disturbance passed successive stations electric circuits were broken and a 

 mark made upon the surface of a smoked glass plate which was moving 

 at a known rate. 



The results, which have not yet been worked out in detail, are gene- 

 rally as follows : — 



1. The velocity of propagation of both normal and transverse vibra- 

 tions is a function- of the initial force creating the disturbance ; or, 

 briefly, the greater the charge of dynamite the greater the velocity. 



2. The velocity of propagation of both normal and transverse vibra- 

 tions is greater between points near to the origin than between points 

 which are distant. Near to the origin this velocity decreases much more 

 rapidly than it does at a distance. 



When writing on this subject, I shall refer to the work done in this 

 direction by Mr. Robert Mallet and by General Abott. 



VI. Experiments to determine the relative motion of two neighhouring 

 points of ground. 



In these expei'iments two stakes were driven in the ground at various 

 distances apart up to about 2 ft. 6 in. ' Sometimes the stakes were so 

 placed that a continuation of the line joining them passed through the 

 origin of the disturbance, and sometimes they were placed at right angles 

 to such a direction. Fixed horizontally upon the head of one of the stakes 

 was a light rigid bar, from the end of which a light index hung vertically. 

 This index at a short distance below its point of suspension, which was a 

 universal joint, was caught by a second universal joint at the end of a bar 

 passing from the second stake. So long as the two heads of the stakes 

 synchronised in their motions, it was assumed that the universal joints at 

 the ends of the bars would keep vertically beneath each other, and the index 

 which they supported would remain perpendicular. If, however, there 

 was a want of synchronism, the lower end of the index would give a 

 multiplied representation of their relative displacements. In all cases it 

 was found that there was a considerable relative motion in the direction 

 of the origin. 



The chief practical value of this experiment was to see how far we 

 are justified in placing two portions of a seismograph upon different 

 stakes. It also shows that a building, although it may be small, may not 

 be moved as a whole, but may suffer considerable racking. 



VII. Experiments on the Production of Earth Currents. 



From near the scene of the explosions a telegraphic communication 

 was established across a deep moat up to a hill where the mechanical' 

 disturbances were practically not observable. Each end of this circuit 

 was put to earth by means of two long crow-bars, and in the circuit on 

 the hill one of Clark's diff'erential galvanometers was arranged. As 

 either of the crow-bars was raised or depressed it was found that the 

 current passing through the galvanometer varied, sometimes being positive 

 and sometimes being negative. At a certain depth, which was found by trial,, 

 the needle of the galvanometer remained at zero, and it was in this way 

 that the adjustment to ' no current ' was made previous to an explosion. 



When the explosion took place, one earth bar being at distances of 

 from 10 to 50 feet, a considerable current was always produced, and the 

 needle of the galvanometer swung with violence until it reached a stop. 

 The direction of swing was, in the few experiments which were made,, 



