Vol.  49.]  GRANITE  IX  THE  GABBRO  OF  THE  CUILLIN  HILLS.  187 
of  the  U.S.  Geological  Survey,  on  the  spherulitic  rocks  of  Obsidian. 
Cliff  in  the  Yellowstone  Park,  and  on  the  similar  rocks  of  Silver 
Cliff,  Custer  County,  Colorado.  The  former  gentleman  has  greatly 
facilitated  my  task  by  supplying  me  with  a  series  of  specimens 
illustrating  the  chief  types  which  he  has  so  well  described. 
The  spherulites  of  the  Cuillin  Hills  sometimes  attain  a  diameter 
of  nearly  f  inches  (60  centimetres)  ;  they  present  examples  of  all 
the  varieties  described  b}r  Messrs.  Iddings  and  Whitman  Cross, 
with  some  forms  not  previously  noticed.  So  far  as  size  goes,  they 
fall  far  short  of  the  spherulites  described  by  the  latter  observer 
from  Silver  Cliff,  which  are  said  to  sometimes  attain  a  diameter  of 
10  feet ;  the  Cuillin-Hills  spherulites,  however,  I  believe,  surpass 
in  size  any  hitherto  described  in  the  British  Isles,  with  the  exception 
of  the  fine  hollow  spherulites  of  Boulay  Bay,  Jersey,  while  they  are 
in  a  much  less  altered  condition  than  these  latter. 
Many  of  the  spherulites  in  question  are  beautiful  examples  of 
what  Mr.  Butley  has  proposed  to  call  ‘  composite  spherulites/ 1 
in  which  globular  spherulites  formed  round  a  number  of  sporadic 
centres  have  been  enclosed  in  other  radially-formed  masses  of  much 
larger  dimensions — they  are,  in  short,  spherulites  within  spherulites. 
The  smaller  spherulites  nearly  all  belong  to  the  ordinary  type 
(the  4  compact  spherulites  ’  of  Iddings),  in  which  microlites,  arranged 
radially,  give  with  crossed  nicols  the  usual  black  cross  or  sometimes 
a  greater  number  than  four  dark  brushes  not  at  right  angles  to  one 
another  (the  so-called  4  pseudo-spherulites  ’). 
Hot  unfrequently,  these  ordinary  spherulites  consist  of  several 
concentric  layers,  the  microlites  of  each  of  which  exhibit  some 
differences  in  colour,  opacity,  or  other  characters.  These  spherulites 
are  found  to  be  involved  in  peculiar  arborescent  growths,  which 
have  been  compared  to  foxes’  tails ;  and  these  start  from  one  or 
sometimes  several  centres,  and  involve  alike  the  ordinary  spherulites 
and  the  phenocrysts  of  the  original  rock,  which  have  escaped  fusion. 
The  bodies  built  up  by  these  arborescent  growths  constitute  the 
4  porous  spherulites  ’  of  Iddings  2  (see  PI.  III.  fig.  7). 
When  the  curious  plume-like  or  brush-like  bodies  are  examined, 
especially  in  polished  surfaces  of  the  spherulites,  which  can  be  viewed 
by  reflected  light,  we  see  that  there  are  two  varieties  of  them,  which 
are  strongly  contrasted  in  colour  (PI.  III.  fig.  7).  In  one  case 
the  colour  is  bluish  grey,  while  in  the  other  it  is  creamy  white. 
Microscopic  sections  of  extreme  thinness  show  that  the  blue-grey 
plumes  are  composed  of  felspar  and  quartz,  through  which  grains 
of  magnetite  are  uniformly  distributed  (PI.  III.  fig.  8).  The 
creamy  white  plumes  are  also  composed  of  felspar  and  quartz,  but 
are  destitute  of  the  scattered  grains  of  magnetite  ;  they  exhibit,  how¬ 
ever,  a  quantity  of  hydrous  brown  oxide  of  iron  lying  between 
the  felspar-fibres  of  the  plumes  (PI.  III.  fig.  9).  Careful  study 
of  very  thin  sections  with  high  powers  shows  that  this  brown  oxide 
1  *  Composite  Spherulites  in  Obsidian,  from  Hot-Springs,  near  Little  Lake, 
California,’  Quart.  Journ.  Geol.  Soc.  vol.  xlvi.  (1890)  pp.  423,  424. 
2  ‘  Obsidian  Cliff,  etc.,’  p.  278. 
