ON SEISMOLOG1CAL INVESTIGATIONS. 



217 



under discussion, naturally adding a good deal both to the work of the year and the 

 printing bill. 



The discussion of the P and S residuals for the five years 1918-1922 has been 

 completed and published in the Geoph. Supp. to the Monthly Notices R.A.S. Copies 

 of the paper will be distributed to the various stations along with the number of the 

 Summary for 1924, January-March. One incidental outcome of the discussion is 

 that there seems to be near A=57°, a minimum frequency of records both in P and S. 

 The following are the ratios of totals (on an arbitrary scale) to the areas of successive 

 zones of the earth's surface, 5° in width : — 



The phenomenon may, of course, be a simple consequence of the particular dis- 

 tributions of epicentres and observatories. The former tend to be near the 

 Philippines, and the latter in Europe, at distances not far from 90° ; and this may 

 cause the rises in the ratio after A=67-5°, which might continue to fall with a more 

 uniform distribution. But the point is worth further examination. 



The corrections to the adopted tables may be represented as follows : — 

 Firstly, if we take the simple maximum of the residuals in P we get 



Corrections to Tables op P. 



It will be seen that besides the well-known negative errors after 85°, there is a 

 considerable error near 35°. But this is accompanied by a curious apparent 

 duplicitj' in the maximum, which is even more striking in the case of S than of P. 

 The results may, therefore, be presented more fully as follows : — - 



Corrections to Adopted Tables for P and S. 



