OP  THE  ISLAND  OE  CAPRAJA. 
138 
Yol.  49.] 
surface  and  along  the  cleavage-cracks,  to  that  in  which  all  hut  a 
small  central  kernel  has  been  4  epidotized.’  Where  the  original 
mineral  has  been  completely  decomposed,  the  epidote  has  a  core  of 
calcinm  carbonate.  Besides  calcium  carbonate,  zeolites  and  opal 
are  almost  invariably  present :  the  former  as  an  exterior  zone  round 
the  epidote,  reminding  one  of  the  green  zone  in  certain  gabbros 
round  olivine  in  contact  with  felspar  ;  the  latter  in  small,  irregular 
patches  in  close  proximity  to  the  epidote,  its  presence  recognizable 
with  certainty  only  after  treating  the  slide  with  hydrate  of  potash 
and  fuchsine.  Much  less  frequent  is  the  occurrence  of  epidote  in 
felspar,  where  it  presents  several  striking  characteristics.  Here  it 
is  always  clear  and  colourless,  and  unaccompanied  by  any  of  the 
other  above-mentioned  products  of  decomposition.  It  is  to  be  found 
only  in  felspar  with  zonal  structure,  and  then  always  in  the  more 
basic  part  towards  the  centre  of  the  crystal.  In  one  instance  the 
epidote  was  seen  to  follow  the  contours  of  one  of  the  interior  basic 
zones,  without  extending  into  the  acid  zone  beyond.  The  felspar, 
beyond  the  presence  of  epidote,  shows  no  signs  of  decomposition ; 
in  fact,  everything  points  towards  the  conclusion  that  this  epidote 
cannot  be  referred  to  augite  as  its  source,  but  is  either  an  original 
inclusion  of  unknown  origin  in,  or  an  alteration-product  of,  the 
felspar,  contemporaneous  with  its  growth.  In  the  latter  case,  it 
might  be  due  to  magmatic  influences  during  consolidation. 
The  Groundmciss  plays  a  very  important  part  in  all  the  Caprajan 
andesites.  In  it  are  always  found  glass,  felspar,  augite,  generally 
magnetite,  and  not  seldom  mica  and  hornblende.  The  relative  pro¬ 
portions  of  glass  and  crystalline  constituents  vary  within  wide  limits. 
The  glass  is  light  brown  to  colourless,  and  may  be  almost  free  from 
microliths  or  very  rich  in  them.  Where  the  brown  varieties  of  glass 
come  into  contact  with  felspar  they  sometimes  lose  their  colour,  and 
become  almost  as  clear  as  the  felspar  along  the  line  of  contact ;  at 
the  same  time  the  microliths  disappear  entirely  or  almost  entirely. 
The  felspar  of  the  gronndmass  is  idiomorphic  and  generally  lath¬ 
shaped.  Crystals  are  single  or  once  twinned,  but  too  small  to  admit 
of  any  determination  of  the  nature  of  the  felspar  from  extinction- 
angles.  The  augite  occurs  in  long,  narrow  prisms,  which  often 
group  themselves  together  by  twos  and  threes.  The  hornblende  has 
already  been  described  as  it  occurs  in  the  gronndmass  (see  p.  138). 
Mica  is  not  infrequently  present,  generally  well  crystallized  in 
minute  hexagonal  flakes. 
Besides  the  above-mentioned  constituents,  there  often  occur  small 
patches  of  a  perfectly  colourless,  isotropic  substance,  which  was  at 
first  mistaken  for  sodalite.  The  excellent  micro-chemical  reactions 
for  sodalite,  proposed  by  Lemberg,1  proved,  however,  that  it  was  not 
1  The  slide  is  treated  with  dilute  nitric  acid  and  a  few  drops  of  silver 
nitrate  in  the  cold  for  ten  to  twenty  minutes,  and  then  washed.  Where 
sodalite  is  present,  a  film  of  silver  chloride  has  been  deposited,  which  soon 
darkens  and  reveals  the  mineral  with  certainty.  I  have  used  this  method 
with  the  greatest  success  on  the  trachytes  of  Ischia  and  the  Phlegrasan  Fields. 
