﻿D 
  MALLET 
  : 
  VOLCANOES 
  OF 
  BARREN 
  ISLAND 
  AND 
  NARCONDAM. 
  

  

  But 
  such 
  lava 
  streams 
  as 
  issued 
  from 
  the 
  main 
  crater 
  (and 
  not 
  from 
  

   lateral 
  vents 
  on 
  the 
  slopes 
  of 
  the 
  mountain) 
  would 
  overflow, 
  or 
  break 
  

   through, 
  the 
  lowest 
  part 
  of 
  the 
  rim, 
  thus 
  tending 
  to 
  neutralise 
  the 
  in- 
  

   equality 
  just 
  mentioned, 
  and 
  the 
  greater 
  resistance 
  that 
  the 
  solidified 
  lava 
  

   would 
  doubtless 
  present 
  to 
  subsequent 
  pluvial 
  denudation, 
  1 
  would 
  be 
  

   another 
  element 
  tending 
  in 
  the 
  same 
  direction. 
  Again, 
  the 
  rain 
  during 
  

   the 
  south-west 
  monsoon 
  must, 
  owing 
  to 
  the 
  influence 
  of 
  the 
  wind, 
  fall 
  

   in 
  a 
  more 
  or 
  less 
  oblique 
  direction, 
  the 
  result 
  being 
  a 
  heavier 
  rainfall, 
  

   and 
  consequently 
  greater 
  denudation, 
  on 
  the 
  windward 
  than 
  on 
  the 
  leeward 
  

   outer 
  declivities 
  of 
  the 
  mountain; 
  the 
  reverse 
  being 
  the 
  case 
  with 
  

   reference 
  to 
  the 
  slopes 
  of 
  the 
  crater 
  itself. 
  The 
  interaction 
  of 
  these, 
  

   and 
  perhaps 
  other, 
  causes, 
  has 
  been 
  such 
  that 
  the 
  resultant 
  does 
  

   not 
  admit 
  of 
  being 
  traced 
  with 
  any 
  certainty. 
  It 
  may 
  be, 
  indeed, 
  

   that 
  the 
  superior 
  height 
  of 
  the 
  crater 
  rim 
  towards 
  the 
  south-east 
  is 
  

   due 
  in 
  some 
  degree 
  to 
  a 
  more 
  than 
  average 
  proportion 
  of 
  lava 
  there, 
  

   and 
  the 
  section 
  exposed 
  along 
  the 
  precipitous 
  cliffs 
  certainly 
  does 
  

   show 
  such 
  rock 
  to 
  be 
  strongly 
  in 
  the 
  ascendant. 
  But 
  any 
  such 
  excess, 
  

   if 
  real, 
  2 
  and 
  due 
  to 
  the 
  causes 
  just 
  mentioned, 
  ought 
  to 
  be 
  accompanied 
  

   by 
  a 
  corresponding 
  excess 
  towards 
  the 
  north-west, 
  where, 
  however, 
  the 
  

   crater 
  rim 
  is 
  lowest. 
  

  

  The 
  marked 
  deficiency 
  in 
  height 
  of 
  the 
  north-western 
  half 
  of 
  the 
  

   circumference 
  must 
  then 
  be 
  ascribed 
  to 
  some 
  other 
  origin. 
  A 
  shifting 
  of 
  

   the 
  axis 
  of 
  eruption 
  would 
  tend 
  to 
  produce 
  the 
  feature 
  in 
  question; 
  

   that 
  is 
  to 
  say, 
  if 
  the 
  axis 
  during 
  the 
  explosive 
  eruption 
  by 
  which 
  the 
  

   ancient 
  crater 
  was 
  in 
  all 
  probability 
  formed, 
  was 
  somewhat 
  to 
  the 
  north- 
  

   west 
  of 
  the 
  axis 
  e 
  isting 
  during 
  the 
  time 
  the 
  original 
  cone 
  was 
  built 
  up, 
  

   the 
  north-western 
  side 
  of 
  the 
  cone 
  would 
  be 
  blown 
  away 
  to 
  a 
  greater 
  

  

  1 
  The 
  very 
  perfect 
  preservation 
  of 
  the 
  extinct 
  scoriae 
  cones 
  of 
  Auvergne 
  has 
  been 
  

   ascribed 
  to 
  the 
  rain 
  sinking 
  into 
  the 
  porous 
  material 
  as 
  fast 
  as 
  it 
  falls, 
  so 
  that 
  there 
  is 
  no 
  

   surface 
  wash. 
  No 
  doubt 
  the 
  ash 
  of 
  Barren 
  Island 
  is 
  preserved 
  in 
  a 
  similar 
  way, 
  but 
  far 
  

   less 
  completely, 
  owing 
  to 
  the 
  heavier 
  rainfall, 
  which 
  at 
  Port 
  Blair, 
  the 
  nearest 
  meteoro- 
  

   logical 
  station, 
  and 
  70 
  miles 
  to 
  the 
  south-west, 
  is 
  118 
  inches 
  per 
  annum. 
  Of 
  this, 
  about 
  

   100 
  inches 
  falls 
  during 
  the 
  six 
  months 
  of 
  the 
  south-west 
  monsoon. 
  

  

  2 
  The 
  concealment 
  of 
  the 
  outcrop 
  of 
  the 
  older 
  rocks 
  elsewhere, 
  by 
  the 
  most 
  recent 
  ash, 
  

   prevents 
  any 
  comparison 
  being 
  made. 
  

  

  ( 
  256 
  ) 
  

  

  