EARTHQUAKE SOUND. 343 



A large number of accounts make no reference as to whether the 

 sound was heard before, during or after the shock, 



d t th sh k ^ ^ e rema u i( i er ^ de larger part record the sound 

 as beginning or occurring just before or a few 

 seconds before the main shock ; a fair number record it as occurring 

 during the shock, and a very few as subsequent to the shock. From 

 this evidence it is clear that the shock was in a great number 

 of cases heard some little time (measurable in seconds) before the 

 shock (or the main shock) was sensibly felt. It is also* clear that it 

 continued during the shock ; whilst the evidence that the sound was 

 sometimes not heard until after the shock may be variously interpreted, 

 but most naturally en the assumption that it was simply not noticed till 

 afterwards. The human ear being unable to record very low-pitched 

 sounds, it is more probable that the varying evidence is due to the 

 imperfect receiver than that there was any real difrerence at different 

 localities in the order of the phenomena. My own personal experience 

 of the sounds accompanying such of the aftershocks as were observed by 

 me is entirely in favour of this conclusion : the sound was never 

 heard after the felt shock, but generally sufficiently before it to arouse 

 expectation for the latter. 



In consideration of the low rate of transmission of sound waves in 

 air, about i mile per second, and in consideration also of the only 

 moderate intensity of the sound even in the epicentral regions (insufficient 

 to break glass windows, as we have seen), there is no other possible 

 conclusion than that the more rapid vibrations capable of creating a 

 sound were transmitted to the air from the ground in the immediate 

 neighbourhood of the observer. Assuming then that then* is sufficient 

 testimony in support of the sound having begun to be h« trd slightly 

 befcre the appearance of the felt shock, we must conclude that these 

 more rapid audible vibrations were communicated to the air from the 

 surrounding ground, being borne as subordinate partial vibrations 

 superposed on the not yet felt preliminary tremors of the direct shock. 

 Wherever the sound was not noticed until during the main shock 

 must suppose that they were not of sufficient intensity, or of sufficiently 



