186 TRANSACTIONS OF SECTION C. 



18 km. and 31^ km. (■eleven and twenty miles). There is some evidence in the 

 case of a distant earthquake of a period approximating to 30 km. per second, 

 which would correspond, -with waves of distortion, to a thickness of 30 km. 

 (nineteen miles). However in the present state of our knowledge of these 

 vibrations such calculations are only of speculative interest. 



There must be numerous surfaces of discontinuity in the earth's crust in 

 addition to that forming its lower limit. Such would be the boundaries 

 between great tracts of granite or granitoid gneiss and the basic rocks that 

 in all probability everywhere underlie them ; the surface dividing gneisses and 

 crystalline schists from unmetamorphosed sediments overlying them unconform- 

 ably ; that between hard Palfeozoic rocks and softer strata of later age ; and 

 the surfaces of massive limestones or sills. Wiechert observed at Gottingen, at 

 the time of the Indian earthquake of April 4, 1,905, small horizontal vibrations, 

 superimposed on the others, with a period of only 1^ seconds. He believed 

 that these were due to horizontal distortional vibrations of the local sand- 

 stone formation with a node at its basal surface. He formd the velocity of 

 similar vibrations at the surface to be 250 m. per second, and thence calculated 

 the depth of the sandstone stratum to be 90 m.-'' No doubt similar correlations 

 of terrestrial vibrations and the structure oi the earth's crust may be made in 

 other cases. 



It deserves consideration, however, as to how far it may be possible to 

 add to our knowledge of the earth's crust by experimental work with a view 

 of the determination of surfaces of discontinuity by their action in reflecting 

 vibrations from artificial explosions, a procedure similar to that by means of 

 which the presence of vessels at distance can b© detected by the reflection of 

 submarine sound waves. The ordinary seismographs are not suited for this 

 purpose; the scale of their record, both of amplitude and of time, is too small 

 for the minute and rapid vibrations which would be expected to reach an 

 instrument situated several miles from an explosion, or to distinguish between 

 direct vibrations and those that may arrive a second or two later after reflexion 

 at a surface of discontinuity. As the cylinder on which the record is made 

 would be only in motion while the experiment was in progress, there would be 

 no difficulty in arranging for a much more rapid movement. At the same time 

 it would be desirable to dispense with any arrangement for damping the swing 

 of the pendulum, which would be imnecessaiy with small and rapid vibrations, 

 and would tend to suppress them. It is possible that it might be better 

 to employ a seismograph which records, like that devised by Galitzin shortly 

 before his death, variations of pressure expressing terrestrial acceleration, instead 

 of one which rec'ords directly the movements of the ground. It would, however, 

 probably be found desirable to substitute for the piezo-electric record of pressure 

 employed by Galitzin a record founded on the effect of pressure in varying the 

 resistance in an electric circuit. This is, in fact, the principle of the micro- 

 phone and most modern telephone receivers, but quantitatively they are very 

 unreliable. This would not matter so much for the present purpose, where the 

 time of transmission is the most important feature in the evidence, but satis- 

 factory results even in this respect appear to be given by Brown's liquid micro- 

 phone, from which the record could be taken, if desired, by means of the 

 reflection of a mirror, attached to the needle of the galvanometer. 



Evidence of the structure of the earth's crust is also afforded by observa- 

 tions on the direction and magnitude of gravitation which have been carried 

 out in considerable detail in India and the United States? — especially in the 

 neighbourhood of great mountain ranges. At the present time the" problem 

 of correlating the variations observed with the underground structure is onlv in 

 an embryonic stage. It is probable that our greatest hope of advancing researches 

 with this object is by detailed work in areas which present no marked oro- 

 graphical features, and where the geological structure is already fairly well 

 ascertained. 



The same remarks apply to the results obtained by magnetic surveys. Apart 

 from the marked effect of masses of magnetite in the immediate neighbourhood 

 of the surface, local magnetic irregularities appear to be mainly determined bv 



-' G/ittinf/en Nachrichten. 1907. pp. ?G7-8. 



