42 



REPORTS ON THE STATE OF SCIENCE. — 1916. 



Table Y. 



It will 'be seen that the main feature of the proposed change in the 

 tables is a positive correction greatest about 20°, followed by a negative 

 correction greatest about 30°. Now, this should be shown by the 

 recoi-ds, and apparently it is. The following examples will perhaps 

 suffice for the present ; a complete discussion would not only be unsuit- 

 able for this report, but requires an expenditure of time which has not 

 yet been found possible, for the reason that records for stations near 

 the epicentre are themselves liable to be used for determining the 

 epicentre, so that errors of the tables may be partly compensated by 

 adjusting the epicentre to destroy them. 



If we are fortunate enough to have two stations, equipped with 

 good instruments and time-determinations, one 20° from the epicentre 

 and the other 30°, and in the same aziviuth, then the relative errors of 

 P above indicated could not be masked. We might alter the absolute 

 errors in the same direction, but the difference would be unchanged. 

 Unfortunately such cases are comparatively rare, and for the present 

 the evidence can only be partially stated. Selected examples are as 

 follows : — 



Table VI. 

 1914 March 14" 20^^ 0" 6^ : 39°-2 N. 139°-8 E. Determined by PulJcovo. 



