ON EARTHQUAKES, 1918-1924. 303 



Aug. 14d. llh. Focal depth +0010. 



Sept. Id. 19h. The first of a series of shocks in North Formosa, others following 

 on Sept. 14, 15, 16, 17, Oct. 14, Dec. 1 and 12, in which 50 people were killed and 

 1,000 houses damaged or destroyed. 



Sept. 4d. 17h. Focal depth +0-080 (exceptionally deep). 



Sept. 22d. 21h. On the evidence of the La Paz observations the epicentre may 

 be at 25-2° N. 46-6° W. with a focal depth +0-045. 



Oct. lid. 14h. The S residuals ot this earthquake near A=90° were seen to be 

 consistently negative with a difference from the tables for S approximately linear in A. 

 Almost simultaneously a letter from Dr. Harold Jeffreys drew attention to Gutenberg's 

 suggestion (in 1914) of an S^PcS wave, which traverses the earth's central core as P, 

 and was readily identified with the phenomenon thus noticed, which has since been 

 denoted by [S]. The anomaly had already been noticed in discussing the iarje Earih- 

 quaJces of 1913: B.A. Seism. Ciee., 1917., but the proper explanation of it was first 

 noticed on this occasion {i.e. in 1926, twelve years after Gutenberg's paper). 



Oct. 17d. 6h. Possibly 18-0'' N. 97-0° E. as on 1919 Sept. 8, but with focal height 

 -0-030. 



Oct. 24d.21h. Focal depth +0-010. 



Nov. 3d. 12h. Focal depth +0-040. 



Nov. lid. 4h. A disastrous earthquake felt over the whole of Chile, ' between 

 Antofogasta (lat. —23°) in the north and Valdivia (lat. —40°) more than a thousand 

 miles to the southof it ' (London Timffs of November 13), 1,800 killed, 35,000 homeless, 

 &c. An investigation by Prof. Baillv Willis suggested an origin ' near the solitary 

 islands of St. FeHx and St. Ambrose'' (say 26-5° S. 80-0° W.), where all the lobsters 

 were killed ; but it seems improbable that the origin could be so far west. There are 

 many observations or [S], not, however, identified as such in the Summary. 



Dec. 6d. 13h. Focal depth +0-020. 



Dec. 19d. 3h. The La Paz observations are inconsistent with those at Mendoza 

 and the former has been preferred. If the latter is preferred, a solution would be 

 Dec. 19d. 2h. 59m. 203. Epicentre 32-9° S. 68-3° W. 



Notes to 1923. 



Jan. 14d. 5h. The macroseismic observations were carefully discussed by Omori 

 who arrived at the position 36° 4' N. 140° 3' E. for the epicentre ; but this does not 

 fit the seismographic results. The earthquake is noteworthy as having occasioned 

 Omori's prediction that ' Tokyo may be assumed to be free in future from the 

 visitation of a violent earthquake like that of 1855 ... as destructive earthquakes 

 do not repeat from one and the same origin, at least not in the course of 1,000 or 1,500 

 years.' The terrible falsification of this hopeful forecast on September 1, 1923, is 

 now matter of history. 



Jan. 22d. 9h. Berkeley gave 41-1° N. 125-5° W., but this does not fit the 

 observations as fully shown in the Summary. 



Feb. 2d. Ih. and 2d. 5h. The residuals for these two shocks are directly com- 

 pared in the Summary and found to be larger than usual ; but the differences appear 

 to be accidental rather than sj'stematic. 



Feb. 19d. 6h. Focal depth +0-060. 



Feb. 24d. 7h. The residuals are divisible into two groups, but no explanation 

 of the difference can be assigned with confidence. The best that suggests itself is 

 A shock at 7h. 34m. 52s. at focal depth +0-020 ; 

 A shock at 7h. 34m. 15s. at focal height —0-020 ; 

 the epicentre being the same in the tM'o cases. 



Mar. 2d. 16h. In the Summary a previous shock from this epicentre on 1922 

 June 27 is erroneously noted. The epicentre on that date is 6-5° N. 126-0° E. 



Mar. 4d. Oh. Perhaps the most puzzling case in our experience. It seems 

 impossible to suggest any solution which does not imply several errors in the observa- 

 tions. A possible alternative is Mar. 4d. Oh. 7m. 5s. Epicentre 16-0° S. 1-0° E. 



Mar. 28d. 20h. On previous occasions a deep focus has been assumed, and the 

 Bame a8.sumption would suit this case also, but the observations are too few to 

 warrant it. 



April 23d. 3h. Focal height —0030. Revised from solution with normal 

 focal depth and epicentre 29-0° N. 124-5° E. 



