﻿Vol. 
  53-] 
  

  

  GREAT 
  JAPANESE 
  EARTHQUAKE 
  OE 
  1891. 
  

  

  13 
  

  

  IV. 
  Distribution 
  oe 
  the 
  After-shocks 
  in 
  Space. 
  

  

  14. 
  Prof. 
  Omori 
  has 
  shown 
  how 
  the 
  after-shocks 
  decline 
  in 
  

   frequency, 
  rapidly 
  at 
  first 
  and 
  more 
  slowly 
  afterwards. 
  This 
  decline 
  

   is 
  accompanied 
  by 
  more 
  or 
  less 
  periodic 
  fluctuations, 
  and 
  is 
  occasion- 
  

   ally 
  interrupted 
  (as 
  on 
  September 
  7th, 
  1892) 
  by 
  strong 
  or 
  violent 
  

   earthquakes, 
  each 
  of 
  which 
  is 
  followed 
  by 
  its 
  own 
  series 
  of 
  after- 
  

   shocks, 
  which 
  may 
  cause 
  a 
  temporary 
  rise 
  in 
  the 
  total 
  number. 
  

  

  That 
  a 
  somewhat 
  similar 
  law 
  governs 
  the 
  area 
  over 
  which 
  shocks 
  

   originate 
  is 
  evident 
  from 
  figs. 
  3-8 
  (pp. 
  5-7). 
  During 
  the 
  first 
  two 
  

   months 
  the 
  epicentres 
  occur 
  over 
  nearly 
  the 
  whole 
  fault-region, 
  but 
  

   afterwards 
  they 
  are 
  distributed 
  over 
  a 
  more 
  limited 
  district, 
  which 
  

   does 
  not, 
  however, 
  continually 
  decrease 
  in 
  size. 
  

  

  In 
  order 
  to 
  ascertain 
  roughly 
  the 
  rate 
  of 
  diminution 
  of 
  the 
  area 
  

   of 
  seismic 
  action, 
  I 
  have 
  made 
  use 
  of 
  a 
  series 
  of 
  monthly 
  maps 
  from 
  

   November 
  1891 
  to 
  December 
  1892, 
  and 
  have 
  in 
  each 
  case 
  determined 
  

   the 
  area 
  included 
  within 
  the 
  curve 
  5, 
  which 
  may 
  be 
  regarded 
  as 
  an 
  

   approximate, 
  though 
  not 
  an 
  accurate, 
  measure 
  of 
  the 
  area 
  in 
  question. 
  

   The 
  results 
  are 
  given 
  in 
  the 
  following 
  table 
  : 
  — 
  

  

  Table 
  III. 
  

  

  Month. 
  

  

  Area 
  within 
  Curve 
  5 
  in 
  

   square 
  miles. 
  

  

  Number 
  of 
  Epicentres. 
  

  

  Nov. 
  1891 
  

  

  1741 
  

  

  935 
  

   930 
  

   633 
  

   572 
  

   728 
  

   543 
  

   444 
  

   432 
  

   473 
  

   546 
  

   622 
  

   727 
  

   467 
  

  

  1139-5 
  

   495-5 
  

   284-5 
  

   1405 
  

   121-5 
  

   123 
  

  

  96 
  

  

  72 
  

  

  84 
  

  

  95 
  

   156 
  

  

  98-5 
  

   103 
  

  

  86 
  

  

  Dec. 
  ,, 
  

  

  Jan. 
  1892 
  

  

  Feb. 
  „ 
  

  

  Mar. 
  ,, 
  

  

  Apr 
  

  

  May 
  ,, 
  

  

  

  July 
  „ 
  

  

  

  Oct. 
  „ 
  

  

  

  Dec. 
  „ 
  

  

  

  The 
  monthly 
  number 
  of 
  after-shocks, 
  already 
  stated 
  in 
  Table 
  I. 
  

   (p. 
  9), 
  is 
  repeated 
  here 
  for 
  convenience 
  of 
  comparison. 
  The 
  figures 
  in 
  

   both 
  columns 
  exhibit 
  a 
  fluctuating 
  decline, 
  but 
  the 
  area 
  of 
  seismic 
  

   action 
  is 
  not 
  proportional 
  to 
  the 
  number 
  of 
  shocks. 
  We 
  should 
  not, 
  

   however, 
  expect 
  this 
  to 
  be 
  the 
  case, 
  any 
  more 
  than 
  we 
  should 
  

   expect 
  the 
  area 
  of 
  the 
  horizontal 
  section 
  of 
  a 
  hill 
  made 
  by 
  the 
  500- 
  

   feet 
  contour 
  to 
  be 
  proportional 
  to 
  the 
  total 
  volume 
  of 
  the 
  hill 
  above 
  

   that 
  plane. 
  

  

  15. 
  Not 
  only 
  does 
  the 
  area 
  of 
  seismic 
  action 
  on 
  the 
  whole 
  decrease 
  

   in 
  size, 
  but 
  the 
  seat 
  of 
  principal 
  activity 
  is 
  subject 
  to 
  almost 
  continual 
  

   change. 
  Thus, 
  the 
  remarkable 
  centre 
  in 
  the 
  rectangle 
  K 
  sinks 
  into 
  

  

  