ON THE PACTS OF EARTHftUAKE PH^ENOMENA. 307 



These results seem to agree pretty well with the American ones as to 

 the degree of consent between eye and ear, or eye and hand, in observa- 

 tional experiments, which gave a probable error for the same observer of 

 about 0"01 of a second ; the mean of the error for both the preceding expe- 

 riments would be 



0-0155 



0-0349 



2)0-0504' 



0-0252 



or nearly double the American error. But the error for observation'with 

 the chronographs, where stopped by hand on seeing the seismoscope wave, 

 would be less, probably not more than the American results on experiments 

 to ascertain difference of longitude by the electric telegraph ; and as two 

 observers were engaged at the same experiments, viz. William and myself 

 each at his own chronograph, the probable error of these experiments may 

 be expressed by V^ x 0-01 or V2 x 0-0252, according to which mean time 

 error be taken. 



We are now in a position to apply the necessary corrections to the first 

 results ascertained for the transit periods both in the sand and in the granite. 



The first correction (see ante), viz. for the time lost by the transit of the 

 galvanic current through the conducting wires used in our experiment, may 

 be neglected ; for assuming Wheatstone's coefficient of 288,000 miles in 1" 

 as its velocity, the total time in the first series is only gss^ooo ^^b ^^ ^ second, 

 and in the second only about one-half of this, both being portions of time 

 far within the limits of possible observation or of instrumentad error. 



The details which have preceded as to the experimental determinations of 

 the probable observational errors, show that any corrections for personal 

 equation are also here a needless refinement. The personal errors indeed 

 were of such a character as to tend to their mutual extermination. 



The only correction then to be made resolves itself into that for the time 

 lost in the production and transmission (through half the length of the mer- 

 eurial trough of the instrument) of the wave in the seismoscope. 



Referring to page 281 we found that the time lost in forming and trans- 

 mitting this wave of the seismoscope, amounted to 0"-065 ; this appears 

 to delay the arrival of the earth-wave or elastic pulse through the me- 

 dium under experiment in its arrival at the instrument. Hence the cor- 

 rection is to convert this time into distance, and add it to the first result 

 given by experiment. 



We found that the gross rate of wave-transit in the sand was 



906-705 feet per second. 



Converting 0"-065 in time into distance, at this rate we have 



906-705 x0"'065 ,„„„,„. • 

 Yfn^ =58-9358 feet; 



and we thus get the final result, 



906-705 + 58-936=965-461 feet per second, 

 the transit rate in the sand at Killiney Bay. 



So for the experiments in the granite we found— for the minimum result 

 in jointed granite — the transit rate, uncorrected, of 

 1220-4'4 feet per second ; 

 and the maximum result in the most solid granite tried was, uncorrected, 

 1559*96 feet per second. 



x2 



