ON SEISMOLOGICAL INVESTIGATIONS. 219 



So far no other hypothesis has been put forward for the explanation of any of 

 these 15+78=93 anomalous eases. While it is not claimed that all of them are 

 convincing in themselves, it is claimed that a large number require some special 

 hypothesis for an adequate solution ; and that when a single hypothesis satisfies them 

 all, the cumulative evidence in favour of it is strong, and can only yield to some 

 alternative which is equally or more successful. 



Near Earthquakes. 



By Dr. Harold Jeffreys. 



When many good seismological stations exist within about 1000 km. of the 

 epicentre of an earthquake their records can be used to give information about the 

 upper layers of the earth's crust. It was discovered by A. Mohorovicic in 1909 that 

 in such cases the records show not only the P and S of ordinary seismology, but also 

 a pair of compressional and distortional waves that have travelled in an upper layer ; 

 their velocities are lower, but their amplitudes greater. Further work by Gutenberg, 

 Conrad and the present writer has shown that three layers are really concerned, which 

 probably correspond to the granitic, basaltic, and ultrabasic layers of geologists. The 

 foci in all cases yet investigated have been in the uppermost, or granitic, layer. Two 

 waves travel in this direct from the focus to the observing station ; these are denoted 

 by P g and S g , and their velocities are about 5-4 and 3-3 km. /see. Others called P* 

 and S* seem to be transmitted down into the intermediate layer, travel along in this, 

 and come up again to the surface. Their velocities in the intermediate layer are 

 about 6-3 and 3-7 km. /sec. Others go right down into the deepest layer. These are 

 the ordinary P and S. Their velocities are 7-8 and 4-35 km. /sec. Thus six distinct 

 pulses are recognisable on the seismograms. 



The times of transmission are linear functions of the epicentral distance, but the 

 constant term is different for every wave, owing to the time spent in the upward and 

 downward journey. The differences indicate that the granitic layer is about 10 km. 

 and the intermediate one about 20 km. thick ; these estimates agree with those 

 made by other means. 



The Jersey and Hereford earthquakes of 1926 have supplied much information 

 in this work. Those used previously were all on the Continent of Europe. 



The velocity of compressional waves in the uppermost layer agrees with that 

 inferred for granite by L. H. Adams and E. D. Williamson from laboratory measures 

 of its compressibility and density. That for the intermediate layer agrees with 

 Adams's and Gibson's experimental value for tachylite, or vitreous basalt ; and that 

 for the lower layer with that of the same authors for dunite, an ultrabasic rock con- 

 sisting mainly of olivine. Holmes has suggested the alternative succession granite- 

 diorite-eclogite. The intermediate layer is not crystalline basalt ; that would give 

 a velocity of about 6-9 km. /sec. 



The observed times of the waves are in accordance with the laws of geometrical 

 optics, but theory and observation both indicate that the amplitudes do not follow 

 these laws, and that diffraction plays an important part. It affects the amplitudes 

 but not the times of arrival. 



The Palestine Earthquake. 



The Palestine earthquake on July 11 must be classed as one of those which excite 

 widespread interest and sympathy rather on account of the nature of the locality 

 than because of special violence. Though undoubtedly disastrous, the intensity of 

 the indications on the Oxford seismograms was far less than that of the China earth- 

 quake on May 22, at a far greater distance. The Acting High Commissioner for 

 Palestine reported on July 18 [The Times of July 19] that in Palestine 200 people had 

 been killed, 356 seriously injured, and 375 slightly injured. At a rough estimate 

 1000 houses were seriously damaged. In Transjordan G8 killed, 102 injured. 



On July 22 and 23 there followed several shocks, one of them considerable, in 

 Persia. 



The British Earthquakes. 



On 1926, August 15, one of the comparatively rare British earthquakes occurred 

 near Hereford and Ludlow. On 1927, January 24, there was an earthquake in 

 Scotland, and on 1927, February 17, there was one in Jersey. The Hereford and 

 Jersey earthquakes have been carefully discussed by Dr. Harold Jeffreys as mentioned 

 above. 



