288 ANNUAL REPORT SMITHSONIAN INSTITUTION, 193 9 



determining the structure of the crust. It is interesting to go back 

 to an early experiment which is in a sense the beginning of our probing 

 of the earth's crust by elastic waves. In 1859 Robert Mallet, an Irish 

 investigator, undertook to measure the velocity of elastic waves caused 

 by blasting in a quarry at Holyhead Island. His methods were crude 

 but carefully carried out, and his distances were short — a little over 

 a mile. He found velocities of from 825 to 1,660 feet per second. In 

 1878 Henry Abbot, an officer of the United States Army, carried out 

 some experiments to determine the velocity of elastic waves caused 

 by gunfire. His distances were greater than those of Mallet — of the 

 order of 13 miles. His deduced velocities were much greater — up to 

 5,900 feet per second. Mallet learned of these results and disputed 

 them. We need not go into details regarding the controversy further 

 than to say that, each being prepared to stand beside his respective 

 determination, the conclusion adopted was that the higher velocities 

 of Abbot's experiment were the result of using gi-eater explosions. In 

 other words: the velocity of elastic waves depends on the intensity 

 of- the generating force. This conclusion was in error. It was con- 

 sidered doubtful at the time and led to further experiments. After 

 some years sufficient data were obtained to establish the true explana- 

 tion vv^hich Ave may now examine. 



Suppose that the upper stratum of the earth transmits elastic 

 waves at a velocity which is less than that for the deeper-lying mate- 

 rial. In this case, deeper means something of the order of 50 to 100 

 feet. The diagram in jSgure 5 shows the approximate path of two 

 rays of the waves propagated from an explosion at the surface. Let 

 us take for granted that the rays are of the form shown and concen- 

 trate on visualizing the speed with which the nose of the advancing 

 wave radiates over the surface. 



If the surface were a calm lake, we could see the water waves 

 radiate in ever widening circles from the shock. Those waves 

 would be gravity waves, not the elastic waves which we are now 

 considering; but the same sort of ever widening circles marks the 

 progress of the energy transmitted by the earth. If the surface 

 were covered with a thin layer of mercury we could actually see the 

 nose of the disturbance spreading out from the source. If we were 

 to put a dish of mercury on the ground at a not too great distance, 

 we could see a segment of the expanding circle pass over it. If we 

 wish to know very exactly when the energy passes regularly spaced 

 points on a line extending outward from the source to a considerable 

 distance, say 20 miles or more, we place sensitive seismographs at 

 those points and have them record on a strip of moving paper on 

 which time marks are simultaneously recorded. In this way we can 

 tell when the initial impulse reaches each station and, if the seismo- 



