ON THE FACTS OF EARTHQUAKE PHENOMENA. 299 



Then 4-867 



4-742 



0-125 total difference, 

 0"-125 

 and — - — = 0"-0625 = the error in defect for one and in excess for the 



other observer, viz. in defect for W. Mallet and in excess for R. Mallet. 



Then, as 0"-0625= error for one revolution of large dial, ?-^2^?^=0"-0052 



= error for one division of large dial, or = one revolution of small dial, and 



^^^=0"-000174=one division of small dial. 



If, instead of the operation adopted, we take the arithmetic mean, 



// 



4-867 

 4-742 

 2)9-607 

 4-8045 

 and then take the difference betvreen each extreme and this mean, 



II II 



4-8670 4-8045 



4-8045 4-7420 



0-0625 0-0625 



the result will be the same. 



The timing of Sharp's chronograph was now proceeded with, and as the 

 whole period to be measured (in the granite) was expected to be much 

 Bhorter than in the Killiney sand, and hence greater accuracy was needed, 

 some weight was added to make this chronograph run faster than on occasion 

 of the experiments there, on principles already explained. 



Sharp's Chronograph (observer's end). 20th Sept. 1850. 

 Five complete revolutions made in each experiment. 



Exp. 1. 5 revolutions made in 21-001 t,t ,, „ 



2. 5 „ „ 21-25 I W'^- M^"^' 



3. 5 „ „ 21-00 r observed 



4. 5 „ „ 2I-25J tf^e clock. 



^- ^ " " 21-801 p , , nyr II . 



6. 5 „ „ 21-50l^°'f''^M^"«t 



7. 5 „ „ 21-45 ^ observed 



8. 5 „ „ 21-75 J tl^e clock. 



8)171-00 

 21-375 



or five revolutions made in 21 "'375, then ^^ '^'^^=4''-275=time ofonerevo- 



4"-275 ^ 



lution of large dial, and ~j2~= 0"-35625=time of one division of large dial,or 



ofone revolution ofsmall dial, and 5-^1^= 0"-01 1875= time of one division 



of small dial, the half or even quarter of which being capable of being read off 



the instrument is sensible to ^!^^^^^=0"-002969, or to about To^^dth of 



one second. 



