ON SEISMOLOGICAL INVESTIGATIONS. 



31 



instruments at Shide, placed close together on separate piers, was given 

 in some detail, and its connection with internal or external temperature 

 was discussed. The Milne-Shaw machine (M-S) was liable to wander 

 much more than the Milne-Burgess (M-B), and the diHerence was 

 provisionally set down to the difference in instrumental construction, 

 seeing that the piers and situations were so closely similar. But the 

 occasion of necessary small repairs to the instruments was taken as an 

 opportunity to interchange their piers; and as a result the M-B now 

 began to wander more than the M-S. To illustrate what happened it 

 will perhaps suffice to give the first harmonics of the daily wanderings, 

 the earlier of which are quoted from the last Eeport: — 



Each result is deduced from the mean of several consecutive, or 

 nearly consecutive, days, for which complete readings are available for 

 both machines. There are some curious points about the behaviour, 

 especially the considerable change of phase in both instruments after 

 the interchange of piers. The changes of sensitiveness * clearly explain 

 a part (even a large part) of the diminution of the coefficient for M-S. 

 But the facts (1) that the M-B coefficient exceeded the M-S after the 

 interchange, and (2) that the difference of phase changed sensibly, seem 

 to show that the difference of behaviour is due as much to the piers as 

 the instruments; and this was specially suggested by a severe rain- 

 storm on September 24-5, which caused, the M-S trace to wander 

 wildly, while leaving the M-B comparatively undisturbed. It is very 

 remarkable that two piers close together in the same building, erected 

 with the intention of being closely similar, should behave in such 

 different ways. After the rainstorm Mr. Bullock carefully examined 

 the foundations of the piers, but without finding anything to explain 

 the difference of behaviour. 



The figures given above show that several points require further 

 investigation before final conclusions can be drawn; but provisionally 

 it would appear: — 



(a) That since two similar piers close together may be disturbed in 

 sensibly, and even seriously, different ways, a locality cannot be 

 judged on the evidence of one test pier alone. If the fauFt lies in the 

 workmanship of one of the Shide piers, there may be an equally 



♦ Allowing for the sensitivity, the ratios of M-S to M-B are 1-5, 8-2. 2-8, and 

 1*5 : then 0-6, after change of piers. 



