54 H. I. JENSEN. 



tells me that during the St. Thomas earthquake of that 

 year, all the ship's compasses ceased to act, only regaining 

 their powers a few hours after. 



Milne, in "Earthquakes," mentions that before the great 

 Japan earthquake in 1855, the owner of a spectacle shop in 

 Asaknsa noticed that his magnet had lost its strength, 

 which it only regained a few hours after the shock. 



There are also a few subsidiary causes, or possible causes 

 worthy of notice. Tlie lessened atmospheric temperature 

 in years of sunspot minimum is quite sufficient to cause a 

 great accumulation of ice at the earth poles. This chaDge 

 of load, increase in cold, decrease in warm regions, may 

 favour volcanic action in tropical regions. 



In conclusion I desire to express my heartfelt gratitude 

 to Professor David for the great assistance he has given 

 me in compiling this paper, by lending me books, giving 

 me access to libraries and suggesting matters, and in bring- 

 ing this paper before the Royal Society. My thanks are 

 also clue to Mr. H. O. Russell for his kind and valuable 

 assistance. 



Explanation of Chart. 



The earthquakes between 1780 and 1902 fall into eleven 

 groups, called a, /?, y, etc., respectively, <* being the most 

 recent. The a group comprises the earthquakes and erup- 

 tions between 1900 and 1902 (inclusive). The fi group 

 includes those around the minimum of 1888-9; the minimum 

 began in 1886 and ended in 1891. The y group 1876—1881; 

 8 1865—1869; € 1852— 1857; (1843—1846; ^1832—1835; 

 1819—1828; i 1810—1813; * 1797—1799; A 1783. 



Any large eruption or earthquake is given a space on the 

 chart, very large ones are given twice as much space as 

 smaller ones, and exceptionally severe seismic phenomena 

 are alloted three times as much. Minor activities, too 

 important to be omitted and yet not important enough to 



