686 



NATURE 



[M«AY 27, 1922 



earthquakes are known to have been accompanied by 

 movement along pre-existing fault-places, in others the 

 origin evidently agrees in position with known faults, 

 and in all of these the distribution of the intensity of 

 disturbance is closely correlated with the faults, being 

 greatest in proximity to them and decreasing as the 

 distance becomes greater. So much is indisputable, 

 yet in spite of a general acceptance of the explanation 

 that the earthquake was a result of the same process 

 which gave rise to the formation of the fault, it must be 

 recognised that the proof is not logically complete, for 

 it might be that the cause and process, which gave 

 rise to the earthquake, were wholly different from, and 

 independent of, those which produced the fault, the 

 only connection being that the weakness, resulting 

 from the fault fracture, served to localise the yielding 

 and so controlled the distribution and intensity of the 

 earthquake. In a study of the Californian earthquake 

 of 1906, where the greatest intensity of disturbance 

 ranged along the line of the San Andreas fault, and was 

 accompanied by considerable displacement and dis- 

 tortion of the surface along the fault-line, I was able to 

 show that the earthquake was due to some cause quite 

 distinct from that which produced the fault, and that 

 neither was the fault the cause of the earthquake, nor 

 the earthquake an incident in the formation of the 

 fault. 



Much more weighty and suggestive evidence is to 

 be derived from some other great earthquakes which 

 have been studied in detail. The conclusion drawn 

 from the Californian earthquake is more fully exempli- 

 fied by the Indian earthquake of 1897. Here there 

 was no single leading fault and zone of maximum 

 intensity of shock, but a complicated network of lines 

 of extreme destructiveness ramifying over an area not 

 much different in area from that of England, and 

 extending right across a series of great tectonic features, 

 across the great monocline of the southern face of the 

 Assam range, across that range itself, across the alluvial 

 plain of the Brahmaputra Valley, the great boundary 

 faults of the Himalayas, and probably even across 

 the main axis of elevation of the range. 



A still more instructive instance is the Charleston 

 earthquake of 1886. There, in a region as devoid of any 

 great structural feature, either of folding or faults, and 

 as little subject to earthquakes, as could be found in 

 our own country, there suddenly occurred a great 

 earthquake, of destructive violence in the central 

 area, and felt over an area measuring about 1500 

 miles across. It was an earthquake of first-class magni- 

 tude, whether we regard the maximum violence of 

 shock, or the extent of area affected, yet there is 

 nothing in the structure of the surface rocks which 

 would suggest its origin being in any tectonic 

 process, and equally nothing which could lead to its 

 classification as volcanic. If we accept the con- 

 clusions of Col. Harboe, regarding the character and 

 extent of earthquake origins, the absence of any 

 connexion between the origin of the earthquake and 

 the tectonics of the surface rocks becomes absolute, 

 for, according to this interpretation, the origin becomes 

 almost coextensive with the seismic area, and the 

 diminution of violence in the outer portions is not 

 due solely to enfeeblement, resulting from the elastic 

 propagation of the earthquake wave, but very largelv 



NO. 2743, VOL. 109] 



to a diminution in magnitude of the originating 

 impulses. 



The interpretation is, I believe, in the main, well 

 founded, and, if it be true that earthquakes of great 

 extent are due to systems of fracture, or analogous dis- 

 turbance, ramifying over, and practically coextensive 

 with, the seismic areas, of which the dimensions in 

 any direction may be measured in hundreds of 

 miles, it becomes more than ever necessary to 

 recognise that earthquake origins cannot be the 

 result of processes and displacements recorded, and 

 indicated by, the tectonics of the surface rocks. The 

 real and ultimate origin must be more deep-seated, 

 and involve either a displacement of, or a change of 

 volume in, the material underlying the outer crust. 



This is no occasion to enter into detail, so I have 

 merely indicated the general character of the studies 

 which have gradually forced me to the conclusion that 

 great earthquakes, and also to a large extent those 

 lesser ones which are commonly classed as tectonic, 

 do not owe their origin to the tectonics of the outer 

 crust, but to processes and changes which take place 

 in the material below it. What these processes may 

 be we cannot know with the certainty which comes 

 from direct observation, for such knowledge as we think 

 we have comes from inference, deduction, and, to 

 some extent, simple assumption, but suggestions have 

 been made which possess a considerable degree of 

 probability. Among these, and especially apposite to 

 present considerations, may be placed Dr. L. L. 

 FermoT's studies of the changes in mineral aggregation 

 which may take place in the solidification of a magma ; 

 he has suggested that the determining factor in deciding 

 the form in which the rock finally solidifies is the inter- 

 relation of pressure and temperature, and has shown 

 that the change of volume, consequent on the change 

 from one mode to another, may amount to more than 

 20 per cent, in extreme cases. Mr. W. H. Goodchild 

 has also studied the subject from another point of 

 view, and suggested that some of the changes, especially 

 the separation of metallic sulphides, take place with 

 great, even explosive, rapidity. 



It is not improbable that, in the material beneath 

 the outer crust, changes of this character are taking 

 place, some slow and gradual, others more rapid and 

 sudden, but all accompanied by a greater or less change 

 of bulk, either of increase or decrease ; and if this be 

 accepted we find an explanation, not only of the forms 

 and origin of earthquakes, but of many other 

 phenomena which are difficult of explanation on any 

 hypothesis of contraction and compression alone. 

 On one hand, slow movements of elevation such as 

 that of the northern Scandinavian region may be 

 attributed to slow and gradual change, involving 

 the whole bulk of large masses ; the lesser earth- 

 quakes may be due to more rapid changes in smaller 

 portions ; the greater to transformation involving a 

 larger bulk of material, and possibly a more abrupt 

 change of combination and density ; while the greatest 

 earthquakes, of first-class magnitude, result from 

 similar changes involving still a larger bulk of material 

 and greater change of bulk. 



To elaborate these considerations forms no part of 

 my purpose ; enough has been said to show that, even 



