546  PROF.  J.  W.  JUDD  ON  COMPOSITE  DYKES  IN  ARRAN.  [YoV.  1893^ 
I.  Analysis  by  Mr.  J.  A.  Schofield,  A.E.C.S.,  of  the  augite-andesite 
taken  from  the  side  of  the  Cir  Mhor  dyke.  The  sulphur  was 
evidently  present  as  iron  disulphide,  which  is  so  frequently  conspi¬ 
cuous  in  these  Tertiary  igneous  rocks.  The  rock  nearly  always 
shows  some  sign  of  alteration,  which  accounts  for  the  water  piesent 
II.  Analysis  of  a  specimen  of  the  quartz-felsite  from  the  central 
part  of  the  dyke.  The  rock  selected  was  a  perfect  4  lithoidite  or 
quartz-felsite,  and,  as  will  he  seen,  the  proportion  both  of  potash 
and  soda  is  high.  This  analysis  was  also  made  by  Mr.  J.  A. 
Schofield.  . 
III.  Analysis  by  Mr.  E.  C.  Thomson,  A.E.C.S.,  of  the  4  pitchstone- 
porphyiy,’  taken  from  the  centre  of  the  dyke.  As  in  most  glassy 
rocks,  the  proportion  of  water  is  very  high ;  but,  seeing  that  more 
than  2  per  cent,  of  the  water  was  lost  at  120°  C.,  it  is  probable  that 
a  part  of  it  must  be  regarded  as  hygroscopic.  The  soda  in  this  rock 
is  in  excess  of  the  potash.  The  numbers  given  are,  in  each  case, 
the  means  of  two  closely  concordant  determinations. 
It  will  be  seen  that,  while  Analysis  I.  corresponds  with  that  of 
normal  augite-andesites,  and  shows  a  very  close  agreement  with  the 
Tertiary  hTorth-of-England  dykes  described  by  Mr.  Teall,  Analyses 
II.  and  III.  indicate  a  magma  of  totally  different  chemical  compo¬ 
sition,  and  one  which  is  of  an  extremely  acid  character. 
The  differences  between  Analyses  II.  and  III.  are  accounted  for 
when  we  remember  that  the  rocks  contain  numerous  crystals  of 
quartz,  felspar,  pyroxene,  and  magnetite  very  irregularly  distributed 
through  them  ;  and  that,  without  elaborate  sampling  of  large  masses 
of  rock,  it  would  be  impossible  to  get  good  average  specimens  for 
analysis.  Moreover,  the  quantity  of  water  in  III.,  as  is  usually 
the  case  with  vitreous  forms  of  rock,  is  abnormally  high.  Consi¬ 
dering  the  two  analyses  together,  we  may  safely  regard  them  as 
representing  varieties  of  the  same  rock,  and  hold  that  II.  differs  from 
III.  only  by  the  devitrification  of  the  glassy  groundmass.  A  very 
marked  feature  is  the  high  percentage  of  soda,  which,  in  III.,  is 
in  excess  of  the  potash.  If,  with  Lagorio,  we  regard  most  of  the 
potash  as  belonging  to  the  glass,  then  we  must  have  a  felspar  present 
very  rich  in  soda.  This  conclusion,  as  we  shall  find,  is  supported 
by  the  microscopical  study  of  the  rock  and  also  by  fezabo  s  flame- 
reactions.  We  thus  see  that  the  chemical  analysis  full)’  supports 
the  conclusion  that  the  rock  is  a  4  quartz-panteEerite  with  a  very 
large  amount  of  glassy  base,  which  may  have  undergone  more  or 
less  devitrification. 
YI.  Microscopical  Characters  oe  the  Eocks  composing  the 
Cir  Mhor  Dyke. 
The  study  of  the  two  varieties  of  rock  of  the  Cir  Mhor  dyke  in 
thin  sections  under  the  microscope  reveals  their  essential  differences 
in  an  even  more  striking  manner  than  does  their  chemical  analysis. 
The  characters  of  the  rocks— crystals  and  groundmass  alike — are 
