ON THE EARTHQUAKE AND VOLCANIC PHENOMENA OF JAPAN, 221 



far to demand serious attention from practical astronomers. There are 

 many reasons for believing that earth jjulsations or undnlations have a 

 connection with the escape of fire-damp, and they do not appear to be 

 wholly unconnected with the behaviour of certain physical instruments. 

 For example, as the result of a long series of observations made with an 

 Oertling and a Bunge balance, it seems that there are times when it 

 would be impossible to carry out any delicate weighing operations. 



The more important results obtained from the observation of these 

 balances were as follows : — 



1. The Oertling, which was a light assay balance, moved more than 

 the Bunge. 



2. It was seldom that either of the balances was absolutely at rest. 



3. Daring a day the pointer of the Oertling usually crept through 

 half a division of the ivory scale. 



4. Although when caused to swing the period of the Oertling was 

 41 seconds, it would sometimes be found performing complete swings 

 with periods varying between 17 and 60 seconds. Slower motions might 

 take 50 minutes. 



5. It was often observed that both balances would start from rest 

 simultaneously and in the same direction. 



6. Periods of disturbance usually occurred with tromometric disturb- 

 ances, but both balances have often been found moving when tremors 

 were not observable, when the weather was calm and the barometer high, 

 while they have been absolutely at rest during a heavy gale and the 

 barometer at 29*2 inches. 



7. The oscillations are not always about the same zero, and the zero 

 for the pointer sometimes changes within a few minutes. 



A detailed account of the above observations is given in the ' Seismo- 

 logical Journal,' vol. i. 



The Earth-waves of Earthquakes. 



From the observations of many who have experienced a large earth- 

 quake we may be certain that at such times the surfaces of alluvial 

 plains have been thrown into a series of undulations. During these dis- 

 turbances, from observations on the behaviour of fluids in vessels the 

 water in ponds, the irregular and erratic swinging of seismographs, and 

 the character of the resulting records, it is also clear that undulatory, 

 wave-like motions have taken place. 



On the occasion of the great earthquake of October 28, 1891, knowing 

 that bracket and conical pendulum seismographs had been tilted, in the 

 Twelfth Report to the Association calculations were given of the 

 maximum slopes of the earth-waves which had caused these movements. 

 Although these calculations may have been interesting on account of 

 their novelty, because any arrangement like a heavy horizontal pendulum 

 when quickly tilted is likely to overswing the point corresponding to that 

 which it would take if the movement had been very slow, serious objec- 

 tions may be raised to the accuracy of the results which were obtained. 

 This consideration led me to devise an angle-measurer in which errors of 

 this description are not likely to occur. It consists of a balance-beam, 

 each arm of which carries a heavy weight so adjusted that the system 

 has but feeble stability. When the stand carrying this is tilted in the 

 plane of the arm, the arm remains horizontal, while a vertical pointer 



