﻿GEOLOGT 
  OF 
  MADAGASCAR. 
  321 
  

  

  sheets 
  of 
  water, 
  with 
  rushes 
  and 
  other 
  aquatic 
  plants 
  around 
  their 
  

   margins. 
  

  

  About 
  90 
  or 
  100 
  miles 
  N.N.E. 
  of 
  Antananarivo, 
  in 
  the 
  province 
  

   of 
  Antsihanaka, 
  it 
  is 
  reported 
  that 
  there 
  are 
  small 
  volcanic 
  craters 
  ; 
  

   but 
  as 
  nothing 
  definite 
  is 
  known 
  respecting 
  them, 
  we 
  can 
  at 
  present 
  

   do 
  no 
  more 
  than 
  merely 
  note 
  their 
  existence. 
  

  

  In 
  addition 
  to 
  the 
  above 
  it 
  may 
  be 
  stated 
  that 
  in 
  some 
  parts 
  of 
  

   the 
  interior 
  of 
  Madagascar 
  (especially, 
  perhaps, 
  in 
  the 
  vicinity 
  of 
  

   the 
  volcanic 
  districts) 
  there 
  are 
  to 
  be 
  found 
  a 
  goodly 
  number 
  of 
  

   circular 
  or 
  oval 
  depressions 
  which 
  are 
  in 
  reality 
  miniature 
  craters. 
  

   These 
  generally 
  occur 
  on 
  the 
  summit 
  or 
  near 
  the 
  summit 
  of 
  the 
  

   low 
  undulating 
  hills 
  so 
  abundant 
  in 
  this 
  part 
  of 
  the 
  island. 
  They 
  

   might 
  easily 
  be 
  mistaken 
  for 
  mere 
  ponds 
  (for 
  they 
  are 
  almost 
  

   always 
  occupied 
  by 
  water 
  or 
  marsh), 
  but 
  on 
  examination 
  cellular 
  

   basaltic 
  lava 
  may 
  frequently 
  be 
  found 
  on 
  their 
  margins. 
  In 
  these 
  

   places 
  volcanic 
  action 
  must 
  have 
  been 
  of 
  the 
  faintest 
  kind, 
  consisting 
  

   chiefly 
  in 
  the 
  discharge 
  of 
  heated 
  vapours. 
  

  

  The 
  largest 
  craters 
  that 
  I 
  have 
  seen 
  in 
  the 
  island 
  are 
  at 
  

   Antongodrahoja, 
  about 
  120 
  miles 
  N.N.W. 
  of 
  Antananarivo. 
  These 
  

   craters, 
  which 
  are 
  in 
  close 
  proximity, 
  occupy 
  the 
  north 
  and 
  north- 
  

   west 
  ends 
  of 
  the 
  hill 
  on 
  which 
  the 
  village 
  of 
  Antongodrahoja 
  

   stands, 
  the 
  remnants 
  of 
  their 
  walls, 
  in 
  fact, 
  forming 
  two 
  of 
  its 
  

   sides, 
  the 
  village 
  being 
  situated 
  between 
  them. 
  There 
  is, 
  at 
  pre- 
  

   sent, 
  no 
  sign 
  of 
  volcanic 
  cone, 
  and 
  only 
  a 
  small 
  arc 
  of 
  the 
  circum- 
  

   ference 
  of 
  the 
  original 
  craters 
  now 
  remains. 
  The 
  cavity 
  to 
  the 
  

   north 
  is 
  a 
  sort 
  of 
  double 
  crater, 
  evidently 
  due 
  to 
  a 
  change 
  in 
  the 
  

   position 
  of 
  the 
  centre 
  of 
  eruption, 
  so 
  that 
  one 
  of 
  the 
  orifices 
  has 
  

   overlapped 
  the 
  other, 
  the 
  remaining 
  arcs 
  of 
  the 
  two 
  crater-rims 
  

   forming 
  a 
  rude 
  resemblance 
  to 
  the 
  figure 
  3. 
  The 
  western 
  volcano 
  

   has 
  been 
  the 
  largest, 
  and, 
  judging 
  from 
  the 
  portion 
  remaining, 
  has 
  

   probably 
  been 
  about 
  three 
  miles 
  in 
  diameter, 
  though 
  no 
  exact 
  

   measurements 
  of 
  it 
  have 
  been 
  taken. 
  The 
  materials 
  poured 
  out 
  

   from 
  these 
  volcanic 
  vents 
  are 
  of 
  a 
  basaltic 
  character*, 
  and 
  appa- 
  

   rently 
  outflows 
  of 
  lava 
  and 
  fragmentary 
  discharges 
  have 
  succeeded 
  

   each 
  other. 
  The 
  basalt 
  assumes 
  a 
  spheroidal 
  character 
  in 
  some 
  

   places, 
  and 
  is 
  now 
  much 
  decayed. 
  Geodes 
  are 
  abundant 
  in 
  some 
  

   parts 
  of 
  the 
  basalt. 
  Just 
  above 
  the 
  waterfall 
  in 
  the 
  northern 
  crater 
  

   they 
  may 
  still 
  be 
  seen 
  in 
  situ, 
  but 
  where 
  the 
  rock 
  has 
  decayed 
  they 
  

   have 
  fallen 
  out, 
  and 
  may 
  be 
  picked 
  up 
  by 
  thousands. 
  The 
  majority 
  

   of 
  them 
  are 
  solid, 
  others 
  are 
  filled 
  with 
  earthy 
  matter 
  ; 
  others, 
  so 
  

   the 
  natives 
  say, 
  contain 
  water, 
  like 
  the 
  enhydros 
  of 
  Brazil; 
  and 
  

   many 
  are 
  hollow, 
  being 
  in 
  various 
  stages 
  of 
  progress 
  towards 
  solidity. 
  

   A 
  few 
  show 
  beautiful 
  agate-markings 
  in 
  cross 
  section. 
  Some, 
  again, 
  

   have 
  whitish 
  horizontal 
  layers 
  deposited 
  against 
  the 
  side 
  of 
  the 
  wall 
  

   that 
  has 
  been 
  lowest 
  in 
  position, 
  showing 
  that 
  the 
  silica 
  has 
  crystal- 
  

   lized 
  in 
  a 
  surplus 
  of 
  water. 
  Many 
  of 
  these 
  hollow 
  geodes 
  are 
  

   remarkable 
  for 
  their 
  beauty, 
  being 
  lined 
  inside 
  with 
  sparkling 
  quartz- 
  

   crystals, 
  some 
  comparatively 
  large 
  and 
  others 
  minute. 
  In 
  many 
  

  

  * 
  Dr. 
  Hatch, 
  in 
  his 
  paper, 
  p. 
  348, 
  describes 
  a 
  specimen 
  of 
  rock 
  from 
  one 
  of 
  

   these 
  craters. 
  

  

  2a2 
  

  

  