ON THE EARTHQUAKE AND VOLCANIC PHENOMENA OF JAPAN. 101^ 



deflected from ^ to |- inch from its true path, and the line, which is often 

 straight, has one or two bow-like bends, suggesting that the column on 

 which the instruments rest has been gradually lifted or warped for a 

 period of, say, five or six hours, and then for the next four or five hours 

 been gradually lowered or gradually returning to its original form. 

 Occasionally this lifting process only takes from fifteen to thirty minutes. 

 The character of the record is hardly that which would be expected as a 

 result of magnetic influences. 



Obseevatioks. 



All my observations have hitherto been made upon a table-like stone 

 column which forms the foundation of a column built some fifteen years 

 ao-o to carry an equatorial telescope. It is 4 feet square, made of 

 carefully squared blocks of stone measuring 4'x2'xl' 3", and rises 

 3' 6" above the floor of one of the rooms in my house which for many 

 years was the Astronomical Observatory of the Imperial University. 

 Unless the shutters of a window are closed for three or four hours the 

 midday situ strikes the south face of the column. A fibre instrument 

 photographically recording N. and S. motion stands near the N.B. 

 corner of the column. On the N.W. corner a second fibre instrument 

 shows B. and W. motion as a spot of light on a scale. Between these 

 two a pair of wire instruments records N.S. and E.W. motions, and a 

 second set of instruments records similar motions on the south side of the 

 column. In between them, from time to time, similar instrunaents have 

 been used in varying positions. 



In regular use there were, therefore, three instruments recording N. 

 and S. motion, and three which recorded E. and W. motion. Observa- 

 tions were commenced on December 18, 1891, and with but few interrup- 

 tions they are being continued. A few of the more important results are 

 as follows : — 



Ti-emors. 



1. On many occasions a stiff" gale has been blowing, and although we 

 should expect tremors if they were a direct eff'ect of the wind, all the 

 instruments have been still : as, for example, on January 15, with W. and 

 S.W. wind; January 18, with N.W. wind ; February 8, with S. wind ; 

 February 24, everything remarkably steady, with N.W. wind; February 27, 

 strong N. wind and very slight tremors ; March 1, N.W. wind ; March 27, 

 W. wind. 



On each of the above days the wind was unusually strong. The 

 I'evolving roof above the column and the shutters rattled, whilst each 

 gust of wind caused my bouse, which is of wood, to shake. When this 

 latter happened, by carefully watching the pointers of an instrument 

 through a magnifying glass, they were seen to tremble, and each 

 trembling, which was momentary, corresponded with the impulses of the 

 wind upon the outside of the building. There was no swing produced in 

 the pointers, but only a temporary vibration along their length. The 

 movements consequent on ' earth tremors ' are distinct and continuous 

 angular displacements, and not the vibration of pointers. 



2. Tremors often occur when the wind is high. They also occur 

 when it is modei-ate, and when it is absolutely calm. The relationship of 



