546 



SCIENCE. 



[N. S. Vol. XXIV. No. 618. 



phenomena that have been made during 

 many years, are, as I take it, to be found 

 in the records of the disturbances that have 

 for many years agitated Japan and in the 

 cataclysmic events which in 1902 visited the 

 Antillean and Gulf regions of America. 

 These additions to knowledge may be 

 briefly suimnarized as follows : 



1. The recognition of the fact that in per- 

 haps the most seismic portion of the earth 's 

 surface— and by application, not unlikely 

 over the whole world — earthquake disturb- 

 ances, especially those of greater magni- 

 tude, are almost invariably preceded, by 

 from one to five or six days, by marked 

 magnetic disturbances in the earth's super- 

 ficial crust. 



2. The close connection existing over vast 

 areas of the earth's surface, as measured 

 on the direct east-and-west line uniting 

 Quetzaltenango, in Guatemala, with Mar- 

 tinique, between the earthquake and vol- 

 canic disturbances of a single or identical 

 period of activity. 



The relations of magnetic disturbances 

 to earthquakes in the Japanese region 

 would seem to be clearly indicated in the 

 records of the earthquakes of Nagoya 

 <^Dec.-Jan., 1893-4), Shonai (Oct., 1894), 

 the great seismic 'tsunanie' of June, 1896, 

 Ugo and Rikuchis (1896), Rikuzen (1900), 

 etc., which have been discussed in the re- 

 ports of the Earthquake Investigation Com- 

 mittee of Japan. While it may be true, 

 as this committee wisely reports, that ' care- 

 ful examination and extensive comparison 

 with magnetic observations abroad, as well 

 as records of earthquakes and pulsatory 

 oscillations,' will be required before any 

 relations between the two phenomena can 

 be considered as definitely established,^ yet 

 the facts are such as to leave little room 

 for doubt in the premises. Indeed, we 

 have the positive statement made by Yama- 



" Publications E. I. C, Tokyo, 1904, p. 81. 



saki, in an address delivered before the 

 geographical section of the University of 

 Vienna, that every major earthquake that 

 visited Japan during the ten years im- 

 mediately preceding 1902, and most of the 

 minor ones, were heralded in advance by 

 magnetic disturbances. If such be the 

 case, and there would appear at this time 

 to be no good reasons* for assuming that 

 the facts are different from what they have 

 been represented to be, then manifestly the 

 cause of earthquakes (and here a special 

 reference is made to those that are referred 

 to as 'tectonic') must be sought for in 

 conditions that have no immediate relation 

 to rock displacements of the time of the 

 earthquake itself— in conditions that may 

 at this time be wholly beyond the field of 

 investigation, and whose expression may 

 be found in some special magnetic or elec- 

 tromagnetic quality of our planet. The 

 remarkable magnetic disturbance which 

 accompanied or followed the cataclysm of 

 Pelee on May 8, 1902, and which in an 

 interval of from less than a minute to two 

 minutes was registered in nearly all the 

 magnetic observatories of the world, from 

 Maryland to Paris, Athens, Honolulu and 

 Zika-wei, in China, may point to some sig- 

 nificance in this connection.^ 



The logical deduction from the premise 

 that has been stated must necessarily be, it 

 seems to me, that, however pointedly a dis- 

 placement along a line of faulting or else- 

 where would seem to indicate the cause of 

 an earthquake, such displacement must 

 rather be looked upon as a consequence of 

 the seismic jar. In other words, the slip- 

 ping or torsion of a terrane did not make 

 the earth-jar, but, on the contrary, it fol- 



* The effect of strain within the earth's crust or 

 mass as of itself initiating magnetic disturbance 

 is not considered in this paper; but the subject 

 has possibly an important place in this inquiry, 

 and its elucidation may help to clear up some of 

 the mystery that attaches to seismo-magnetism. 



