110 REPORT — 1892. 



such tremors to winds at a distance, and to barometrical gradients, have 

 been discussed in previous Reports. Sometimes the tremors are only 

 shown on the instruments recording N. and S. motion, sometimes only on 

 those which record E. and W. motion, and sometimes they are marked 

 on all the instruments. 



8. One of the most important results obtained from the new method 

 of observation is that the so-called earth tremors do not appear to be 

 movements resulting from a general vibration in the soil, such as might, 

 for example, be caused by the falling of heavy weights. With a tremo- 

 meter of the ordinary pendulum type a simple swinging, corresponding 

 to the period of the pendulum, is only seen ; and although it has been 

 observed that the swing is not always uniformly to the right and left of 

 the same point, this swinging might still represent a cumulative effect of 

 minute vibrations. With the conical pendulum arrangement, the inertia 

 of which is extremely small, although they sometimes appear to be 

 moving with their natui-al period, it is oftener that their vibrations 

 appear to be forced. By watching the spot of light as it fitfully swings 

 from side to side, sometimes quickly and sometimes slowly, during a 

 tremor storm, it is difficult to avoid the conclusion that the movements 

 are produced by the tilting of the column. For example, on the night of 

 February 14, when the barometer was low (29'6) and but little wind, all 

 the E. and W. instruments were moving, while the N. and S. were still. 

 The E. and W. light from time to time swung between 32 and 39 (i divi- 

 sion =2"5 mm.), and it seemed as if at irregular intervals, which did 

 not correspond with gusts of wind, that the east side of the column had 

 been lifted. On February 16, in the afternoon, whilst a strong N.W. 

 wind was blowing for some time, the E.W. spot of light rested at 30, 

 then it would move to 34, but return to 30. Five minutes later it stood 

 at 35, but now and then it would move quickly to 40, and return to 30. 

 It did not swing evenly on either side of any point. Similar observations 

 have been made on other days. With the N. and S. recorder these 

 effects, which are apparently due to tilting, have only been observed 

 once or twice, and they were not well pronounced. It is probable that 

 these movements may be due to the same causes, which at the time of 

 very heavy winds I have seen to cause the bubble of a delicate level to 

 pulsate. One result of these observations is that, for the present, at least, 

 I can no longer regard earth tremors as movements which correspond in 

 ■character to vibrations like those which may be produced by a steam- 

 hammer or a railway train — on which I have often experimented — bat 

 rather that they ai'e wave-like pulsations, irregular in period, and 

 producing changes in level. 



They may possibly be the result of sudden variations in barometric 

 pressure over large areas. 



4. During a tremor storm, which sometimes lasts for twenty-four 

 hours, the range of motion is very variable, the intervals between periods 

 of maximum movement being sometimes ten or fifteen minutes. If these 

 motions are due to tilting, this means that the slopes of the wave-like 

 pulsations are not equally steep. 



5. The photographic records, which refer to N. and S. motions — 

 whether there are tremors or whether tremors do not exist — by the 

 deviation of the spot of light, sometimes to the right or sometimes to the 

 left, give evidence of angular motion of the mirror, which may be ex- 

 plained on the assumption of tilting of the ground, warping of the column 



