ON SEISMOLOGICAL INVESTIGATIONS. 



47 



Earthquake Frequency. — As it is possible that an entry which only 

 refers to one station and does not appear to have been noticed in Europe 

 may not have had a seismic origin, in the summation of the above lists 

 such entries have been omitted. Adopting this precaution, the number 

 of earthquake records obtained at the different stations are as follows : — 



Bidston, 33 or 36 ; Shide, 31 or 33 ; Kew, 26 ; Edinburgh, 21.' 



Earthquake Duration. — In summing up the total number of minutes 

 during wlaich the pendulums have been moved, only the fourteen earth- 

 quakes are considered which were recorded or might have been recorded 

 at the four stations. The results in minutes are as follows : — Bidston, 919 ; 

 Shide, 887 ; Edinburgh, 825 ; Kew, 761. 



Accuracy in the Observation of Times of Commencements. — The greatest 

 possible difference in time we should consider likely to exist between the 

 commencement of movement for a given eai'thquake fit two stations would 

 be for disturbances travelling in a northerly or southerly direction between 

 Shide or Kew and Edinburgh, and this could not be expected to exceed 

 five minutes. Between Shide and Kew there might be a difference of 

 one minute, whilst between Bidston and the remaining stations the 

 differences should not exceed two and a half minutes. In the columns 

 relating to these differences the zero indicates the station at which motion 

 was first recorded. The minutes which elapsed before the same dis- 

 turbance was noted at the remaining stations are iiidicated by numerals 

 to the right or left of the zero. 



A minus sign following one of these numerals indicates that the time 

 interval exceeds the expected interval, whilst a plus sign indicates that 

 the numeral is a possible quantity. For the second entry the four minus 

 signs indicate that there are not even two entries which are comparable. 

 In the third entry for February 15, Edinburgh and Bidston, like 

 Edinburgh and Kew, are possible figures, and therefore these three 

 stations are credited with a plus. 



' 74 per cent, of the Shide records are corQmoii to Kew, and 88 per cent, of the 

 Kew records are common to Shide. See pp. 42, 4 d. 



