354 ANNUAL REPORT SMITHSONIAN INSTITUTION, 19 31 



records registered at Jena and Grottingen, respectively — ^two German 

 stations about 80 miles apart — of an earthquake which occurred at 

 Kansu, China, about 4,600 miles from the recording stations. One 

 minute of record line is indicated at the beginning of the record and 

 successive minutes can be seen on the original records as breaks in 

 the line, a few of which are preserved even in the reproduction. 

 The difference in time of arrival of S and P is here about 8 minutes 

 and 50 seconds. To obtain a record such as this a line recording 

 speed of from half an inch to an inch and a half per minute would 

 ordinarily be employed, depending on the seismograph used and the 

 purpose of the registration. Where S and P arrive within a second 

 or so of each other, as in the case of a blast or explosion, it is cus- 

 tomary to have a line recording speed of 6 or 7 inches per second. 

 Such a short-distance but spread out record would resemble those 

 registered at Gottingen and Jena for the Kansu quake except that 

 the serrations would be more numerous. 



The second group of records is also interesting. It shows the 

 seismograms registered at Jena for three earthquakes, each of which 

 originated in Kamchatka — about 5,300 miles distant — in October, 

 1920, June, 1924, and July, 1924, respectively. This shows how 

 strikingly alike are the records of the same instrument for earth- 

 quakes originating at the same epicenter. 



A friend of mine, a devotee of bridge, sometimes mutters a brief 

 prayer in the words, " through strength and up to weakness." Let 

 us repeat this solemn incantation now in order that you may be 

 properly impressed with the strength of that part of seismology 

 which has become so well established that it now forms the strong 

 supporting column of the seismological attack on the problem of the 

 internal structure of the earth. If, at the close of this address, you 

 feel that the van of that attack is occupying ground which it may 

 not be able to hold indefinitely, do not forget that all will not be 

 lost in any event. We are sure of some things — and rather interest- 

 ing and somewhat surprising things they are, too. 



We know, from long experience, that if an earthquake takes place 

 as a sharp, well-defined, single shock, it will be registered at stations 

 which are more than TOO miles and less than 7,000 miles distant in 

 such a manner that the arrival times of P and S can be definitely 

 determined, the difference computed, and the distance from station 

 to epicenter (that point on the surface vertically above the focus) 

 read off from an empirical table or its graph to within 25 or 50 miles. 

 F'urthermore, we can, if the stations are equipped with proper appa- 

 ratus for recording absolute time, determine the time at the epi- 

 center so accurately that the values derived from the records of 

 stations at distances within the above limits will agree within a few 



