Proceed in </s of the Asiatic Society. [August, 



points, will reach the observer's ear at one instant of time, so that 

 the concussions, though separately imperceptible, form, by their cu- 

 mulative effect, a single perceptible concussion, producing the sound 

 as of a gun. It is possible to imagine the occurrence, once in a 

 way, of this exact arrangement of so many different waves, but it 

 is a concurrent arrangement of so many elements as to form, even 

 in a single occurrence, a wonderful coincidence. But that the same 

 coincidence should occur, with respect to the same observer, over 

 and over again at intervals of ten minutes or so, during a single 

 night, is to me quite inconceivable ; and I cannot at present accept 

 a theory which requires me to believe in the frequently repeated 

 occurrence of such an extremely unlikely event. 



The chief argument in support of the surf theorj^ lies in the 

 allegation that the sounds are heard most frequently in a lull fol- 

 lowing a storm, when the waves might be expected to be loudest. 

 But this is a fact somewhat vaguely stated, and without more 

 extended and more exact observation, it cannot be accepted as the 

 basis of any conclusion. 



Another remark I would make on the surf theory is this. We 

 see the production of the noise depends upon the existence of cer- 

 tain conditions as to the comparative distances from the observer's 

 ear of the various points where the waves break. Now to two 

 observers distant from each other by even ten miles, these distan- 

 ces are necessarily quite different, and the same series of waves 

 which combine to produce a perceptible sound upon the ear of one 

 observer, cannot so combine with reference to the other observer. 

 The various concussions will not meet at his ear, but will be scat- 

 tered over a short space of time, and thus be dissipated. 



If therefore we could find as a matter of fact that the sounds 

 were simultaneously perceived by two distant observers, we could 

 deduce from that fact the conclusion that they are not produced by 

 the breaking of waves, or in fact by any other causo which is not 

 strictly confined to one spot, but depends for its effect upon accu- 

 mulation from a number of partial causes (such as the breaking of 

 different waves, or different parts of the same wave) spread over 

 a certain extent of space. The breaking of a wavo two or three 

 miles long, might be conceived so to take place as to produce at 



