ON THK EARTHQUAKE PHENOMENA OP JAPAN. 203 



IV. An endeavour to find out the relative extent of motions and variation 

 in direction of an earthquake in passing over a limited area, the contour 

 and geological structure of ivhich is irregular. 



To work out this problem I have distributed six similar seismometers 

 on the hills and in the valleys near my house. Since being established 

 only four earthquakes have been recorded. Until a number of shocks 

 have been felt I can hai'dly speak of what the result of this investigation 

 may be. 



V. The carrying out of a long series of experiments on artificial earth- 

 q_iMhes in the alluvium of the Tolcio plane. 



These experiments were carried out jointly with my colleague Mr. 

 Thomas Gray. 



These earthquakes were produced by allowing a heavy iron ball to 

 fall from various iieights up to 35 feet. The results obtained were as 

 follows : — 



1. A partial determination, of the eifects of cuttings (like a deep 

 pond) and hills upon the transmission of vibrations. Small hills seemed 

 to produce but little effect in stopping transverse vibrations, but a pond 

 to a certain extent cut off both transverse and normal vibrations. 



2. A complete graphical separation of normal from transverse vibra- 

 tions. 



3. A determination of the relative amplitudes of normal and trans- 

 verse vibrations as observed at points differently situated with regard to 

 the origin of the shock. It appeared that although near to the origin 

 the amplitude of the normal vibrations was greater than that of the 

 transverse ones ; as we made observations at more distant stations these 

 normal motions diminished more rapidly than the latter. Roughly 

 speaking, the amplitude of the normal vibrations was inversely as the 

 distance from the origin of the shock. The transverse vibrations 

 diminished in amplitude more slowly. 



4. There appeared to be usually about six vibrations per second. 

 The normal vibrations were the more rapid. 



5. The average velocity of the normal vibrations was 438 feet per 

 second, whilst that of the transversal movements was only 357 feet. 



6. At a distant station (250 feet) four or five dissimilar vibrations 

 would be first recorded, and then the same four or five vibrations would 

 be repeated in the same order as first recorded. At times this cyclic- 

 like action was very distinct. 



7. The experimental determination (by bending and twisting) of the 

 elastic moduli, &c., of several common Japanese rocks. These experi- 

 ments were performed in conjunction with my colleague, Mr. Thomas 

 Gray, in the physical laboratory of the Imperial College of Engineering. 



In addition to the work which I have mentioned as being before me 

 for the coming seismic season, I shall endeavour to increase the stations 

 where time-observations have been made from two to three or four, at all 

 of which telegraphic signals can be received. I do this, firstly, because 

 I find from experience the important part taken by time-observations in 

 the interpretation of the best of our records ; and, secondly, because 

 they may perhaps be the means of showing us the true relations existing 

 between normal and transversal movements. 



