Vol.  49.]  PROP.  J.  W.  JUDD  ON  COMPOSITE  DYKES  IN  ARRAN.  541 
But  occasionally,  as  I  have  shown  to  be  the  case  at  Beinn  Hiant, 
these  rocks  may  contain  much  glass,  and  they  then  pass  into  the 
1  vitrophyric 9  types,1  the  4  pitchstone-porphyries  ’  of  older  authors. 
There  is  one  characteristic  which  all  these  rocks  present,  namely, 
a  tendency  for  the  glass — which  is  dark  in  colour  and  sometimes 
approaches  to  tachylyte  in  the  abundance  of  its  dusty  particles  and 
skeleton-crystals  of  magnetite,  as  well  as  in  its  forked  microlites  of 
felspar — to  separate  from  the  crystalline  portion  of  the  rock.  This 
separation  may  take  place  in  three  different  ways.  Sometimes  the 
glass  tends  to  form  a  selvage  at  the  sides  of  the  dyke,  or  to 
accumulate  at  its  centre.  At  other  times,  as  is  so  well  illus¬ 
trated  in  the  account  given  of  the  Eskdale  dyke  in  Dumfries¬ 
shire,  by  the  officers  of  the  Geological  Survey  (Sir  Archibald  Geikie, 
Mr.  B.  hi.  Peach,  and  Mr.  J.  S.  Grant  Wilson),2  the  glass  may  se¬ 
gregate  in  more  or  less  irregular  patches  or  bands.  Again,  when 
gas-cavities  are  formed  in  these  rocks,  there  is  a  great  tendency  for 
the  glass  to  ooze  out  from  between  the  crystals,  and  to  fill  these 
cavities,  as  pointed  out  by  Mr.  Teall  in  the  Tynemouth  dyke,  and 
by  myself  in  the  Beinn  Hiant  rocks.3 
The  other  type  of  rock  is  represented  by  the  various  4  pitch- 
stones  ’  of  the  Western  Isles  of  Scotland.  These  rocks  are  of  much 
more  acid  composition  than  the  augite-andesites  already  described. 
Their  silica-percentage  in  the  varieties  containing  much  free  quartz 
ranges  from  70  to  75,  and  in  those  with  little  or  no  free  quartz,  but 
containing  more  or  less  felspar,  from  65  to  70. 
There  is  considerable  difficulty  in  defining  the  place  of  these 
vitreous  rocks  according  to  modern  systems  of  petrographical  nomen¬ 
clature.  This  arises  from  the  circumstance  that  all  these  systems 
of  classification  take  note  of  the  crystalline  minerals,  while  they  to 
a  great  extent  ignore  the  vitreous  groundmass  which,  in  the  rocks 
under  consideration,  usually  forms  by  far  the  larger  proportion  of 
the  mass.  It  i's  the  glassy  basis  which  is  so  characteristic  of  all  the 
rocks  of  the  type  we  are  now  considering,  and  it  is  the  glassy  basis 
which  retains  its  uniformity  of  character,  while  great  variations  are 
seen  in  the  quantity  and  nature  of  the  crystallized  minerals  that 
occur  as  porphyritic  constituents. 
The  glassy  basis,  with  the  minerals  that  have  separated  from  it 
(including  the  dust-like  crystallites,  with  their  ‘  courts  of  crystal¬ 
lization  ’  and  their  fern-like  or  star-like  groups  of  hornblende- 
microlites),  is  well  known  to  petrographers  by  the  descriptions  and 
figures  published  by  Zirkel,4  Allport,5  Teall,6  and  others.  I  shall 
1  Quart.  Journ.  Geol.  Soc.  vol.  xlvi.  (1890)  p.  378. 
2  Proc.  Roy.  Phys.  Soc.  Edinb.  vol.  v.  (1878-80)  pp.  219-254;  and  Trans. 
Roy.  Soc.  Edinb.  vol.  xxxv.  (1888)  pp.  40-44. 
3  Geol.  Mag.  for  1889,  pp.  481-483 ;  and  Quart.  Journ.  Geol.  Soc.  vol.  xlvi. 
(1890)  p.  379. 
4  Zeitschr.  d.  Deutsch.  geol.  Gesellsch.  vol.  xxiii.  (1871)  p.  1. 
5  Geol.  Mag.  for  1872,  pp.  1-10  &  536-545 ;  also  ibid.  1881,  p.  438. 
6  ‘  British  Petrography,’  1888,  pp.  344-347. 
