C. Dcwison — British Earthquakes of 1889. 311 



not possible to draw with any accuracy the isoseismal of intensity 

 IV. As some of these places are not very distant from the liue 

 which I have indicated on the map as the boundary of the disturbed 

 area, we may conclude that that boundary represents an isoseismal 

 of intensity less than IV., but yet greater than III. This area, as 

 before remarked, is nearly circular, and it is almost exactly 

 concentric with the isoseismal of intensity V. 



Sound-Phenomena. — Eecords of sound-phenomena come from 49 

 out of the 156 places where the shock is known to have been felt. 

 These places are indicated as before by small crosses ( + ) drawn 

 through the spots representing them. 



Referring to the map, it will be seen that, with four exceptions, 

 all the places where sounds are said to have been heard lie within 

 an area which is approximately circular. It is of course possible 

 that, outside this area, the sounds might have been heard by 

 observers who were listening attentively or who happened to be 

 more sensitive to such vibrations. This may have been the case 

 in the fovu" exceptions just referred to, but it should be noted that, 

 at each of these places, we have only one record of sound-phe- 

 nomena, while at all the other places there are either several 

 independent accounts, or else the description is so clear as to be 

 quite free from doubt. The following are the accounts from these 

 four places : — 



Bolton-by-Bowland : a rumbling noise. 



Longridge : a noise as of a sough of wind. 



Greenhalgh : an underneath rushing rumbling noise. 



Marbury : a dull rumbling noise. 



"Whilst, then, the characteristic earthquake-sounds may have been 

 heard outside the curve marked (by a dotted line) upon the map, 

 I believe that that curve includes all the places where the sounds 

 were distinctly and certainly heard. 



Assuming this to be the case, two facts are at once evident from 

 an inspection of the map. 



(1) The sounds were not heard so far as the vibrations of longer 

 period were felt. They were confined to a nearly circular district, 

 29 miles in diameter, and about 636 square miles in area. 



(2) The sound-curve is not concentric with the two isoseismal 

 curves drawn upon the map. At its northern limit, it nearly 

 touches the isoseismal of intensity V. ; but, at the other end, it 

 passes about six miles to the south-west of this line. Its centre is 

 about 3^ miles S.S.E. of the epicentrum ; but this determination 

 cannot be regarded as so accurate as in the case of the Edinburgh 

 earthquake, for it is not checked by a knowledge of places where 

 the sound was certainly not heard. 



This second conclusion — the excentricity of the sound-area — is 



also borne out by the following table. The earthquake-sounds are 



recorded as having been heard : — 



"Within the isoseismal of intensity V. at 33 out of 72 places, or in 46 p.c. of the whole. 



Between this isoseismal and the sound- 

 curve at 12 „ 25 „ 48 „ 



Between the sound-curve and the 



boundary of the disturbed area at 4 „ 41 „ 10 „ 



