NOVEMBEB 2, 1906.] 



SCIENCE. 



547 



lowed in obedience to an impact already 

 delivered, and naturally along lines of pre- 

 determined weakness. Indeed, there are 

 many reasons for believing— bnt the facts 

 can not be discussed in this place— that ex- 

 tensive slips along breakage-lines or lines 

 of fracture, in any way correspondent to 

 the long line of displacement which marked 

 the late California earthquake, could not 

 have been initiated without a precedent jar 

 or impact; and still less could there be 

 initiating displacement along a line of con- 

 tact such (e. g., between the looser ma- 

 terials of the coastal plain of the eastern 

 United States and the compact sub- Appa- 

 lachian border-rock) as had been advanced 

 to explain the Carolina earthquake of 1886. 



The extent of field which may be covered 

 by interrelated volcanic and seismic dis- 

 turbances of one period of activity, as is 

 evidenced by the Antillean events of the 

 year 1902, bears directly upon the question 

 of the causation of certain earthquakes 

 which have hitherto been thought to be of 

 a purely tectonic character. The facts con- 

 nected with the Antillean disturbances are 

 briefly : 



The destructive earthquake of Quetzal- 

 tenango, in Guatemala, on April 17-18, at 

 almost precisely the time when Pelee first 

 seriously manifested its new activity; the 

 renewal of activity, immediately after the 

 earthquake, and at a distance of nearly 

 200 miles, of Izalco, in Salvador, a volcano 

 whose energies had calmed down for a 

 number of years, but which was in full 

 activity on May 10, two days after the 

 Pelee cataclysm; the eruption on May 7, 

 of the Soufriere, in St. Vincent; the cata- 

 clysm on May 8, of Pelee, followed, as in 

 the case of the Soufriere, with violent dis- 

 turbances extending into September or Oc- 

 tober; the eruption on October 24 (and 

 continuing to Nov. 15) of Santa Maria, in 

 Guatemala, a volcano situated close to the 



seismic field of Quetzaltenango, and for 

 which there is no recorded previous erup- 

 tion. The relation of these facts, it seems 

 to me, is so conclusive that one need hardly 

 discuss the probability of another interpre- 

 tation being found for them; and it was 

 not without reason, therefore, that Milne 

 early advanced the view that the April 

 earthquake of Quetzaltenango was the real 

 initiator or instigator of the series of dual 

 disturbances that followed rapidly upon it. 

 The lid was taken off the boiling pot, and 

 the pot exploded. Whether or not one 

 should extend the relation of disturbances 

 so as to include the earlier earthquake 

 which in January of the same year wrecked 

 a large part of the town of Chilpancingo, 

 in southern Mexico, and the reawakening 

 of Colima in February and March of the 

 year following (1903), does not materially 

 affect the problem, as the distance separa- 

 ting Martinique from Quetzaltenango is al- 

 ready so great as to fully satisfy the broad 

 deduction which it is the aim of this paper 

 to present. Owing, perhaps, to the fact 

 that these disturbances were developed in 

 what might be termed a single region, and 

 in a region that is not familiar to us in 

 the sense that parts of the world nearer to 

 our homes are, the geologist is not apt to be 

 impressed with the magnitude of the dis- 

 tance that separated them ; it ib, therefore, 

 proper to state that on the map of the con- 

 tinent of North America it would be meas- 

 ured by the line uniting Galveston with 

 Cape Churchill, on Hudson Bay, or that 

 uniting San Francisco with the volcano of 

 Iliamna, on Cook Inlet, Alaska. 



Applying the test of distance in a pos- 

 sible relation touching the history of other 

 (so-called 'independent') earthquakes, we 

 find some not wholly uninteresting results. 

 Thus, the great earthquake of Lisbon 

 (1755), seemingly the most destructive of 

 all the seismic disturbances which history 



