72 



REPORTS ON THE STATE OF SriENCB. — 1915 



on tlie construction of the tables in use requires very careful considera- 

 tion, and is an additional reuson for deferring the proposal of definitive 

 corrections. Indeed, it seems probable that these cannot be made without 

 a somewhat extensive study of the traces themselves in addition to the 

 collation of the times published by the various observatories. 



There is one furthf^r point to be considered — the depth of the focus 

 below the surface. If tliis be at E (see page 73) and CEc be the chord 

 perpendicular to the radivis OEK, then all other chords through E are 



E-W: t= ground to E 



^ \nW^AAAi \/\A^^^^'^^'\/ 



VERTICAL t- ground up 



Eskdalemu'.r, June 25, 1914. 



longer than cC, and occur in pairs such as aEA and &EB. Waves travel- 

 ling from E by either the path Ea or EB woidd on reflection travel by 

 consecutive step? all equal to oA or hB. Hence, cC represents the mini- 

 mum step for waves emanating from E, and is larger as E falls further 

 below the surface. It seems possible that if E is too near the suiface 

 (or perhaps above it) these nearly tangential reflected waves cannot 

 occur, and we get no ' polychord ' or Y phenomenon. This may explain 

 why it has only occasionally attracted notice. Both the epicentres. 



