188 The East Anglian Earthquake.
and 152), the former giving 9h. 20 m. and the latter 9 h. 17
or 18m.96
The accuracy necessary for determining the epicentrum by
means of time-records can in fact be only attained by care-
fully regulated clocks, of which the error can be frequently
determined by electric comparison with some standard time-
piece or by the transit instrument as used in ordinary
astronomical work. Even in such cases a slight error may
be introduced in the absence of seismograph records, owing to
different phases of the same shock being timed at different
stations, instead of the same phase of the earthquake-motion.
From the requirements thus shown to be necessary for exact
observation in this branch of seismology, it is evident that it
would be useless to make use of the present records for
determining the origin of the disturbance with any precision;
but it will nevertheless be of interest to give these observations,
since a certain amount of information can be derived from
them.97
The following list contains all the records which appear to
be trustworthy; mere estimates as to the probable time or
statements that the shock occurred "about" a certain time
have, except in special instances, been neglected:—
1. Colchester......9 h. 18 m. Mr. Hill's large regulator (p. 46).
2. Do.......9 h. 20m. Mr. Aberdein's 3 clocks (p. 49).
3. Do.......9 h. 20 m. (nearly). Mr. W. Marriage's 3 clocks (p. 52)
4. Do.......9h. 20m. Rey. Dr. Manning, St. Leonard's Rectory
5. Birch..........9 h. 20 m. (precisely). Mr. J. Wilkins (p. 58). [(p. 52).
6. Gt. Wigborough. .9 h. 17m. Rev. F. Watson (p. 72).
7. West Mersea___9 h. 18 m. Observed at White Hart Inn (p. 75).
8. Do. ___9h. 15 m. Observed in School-room (p. 76).
96 The Kew record is erroneously printed in the text (p. 152) as
9 h. 17 m. 18 s., instead of 9 h. 17 or 18 m. Mr. G. M. Whipple, the
Superintendent of the Observatory, informs me that neither the Green-
wich magnetograph nor their own is adapted for recording the time of
the shock with the necessary degree of accuracy for Seismological treat-
ment. The time-scale of the Greenwich instrument is 14 mm., and that
of the Kew instrument 16 mm. to the hour, so that the difference of 2 m.
between the records is only about 1/2 mm. on the scales, while the spot of
light is more than a millimetre wide in both instruments.
97 The methods of employing time-data for determining the origin will
be found fully described in Milne's papers (Trans. Seism. Soc. Japan,
vol. ii., p. 69 ; vol. iii., p. 132; and vol. vii., pp. 1 et seq.).