540  PEOF.  J.  W.  JTT)D  ON  COMPOSITE  DYKES  IN  AEEAN.  [Aov.  1 893, 
the  crystalline  and  vitreous  portions  of  the  magma  has  led  to  the 
formation  of  rocks  ranging  in  silica-percentage  from  52*7  to  65*8, 
and  differing  in  a  very  marked  manner  in  texture  and  appearance. 
In  the  fragment  of  a  lava-stream  which  forms  the  Sgur  of  Eigg 
we  have  only  the  more  acid  type  of  rock  represented. 
The  reason  why  such  small  relics  of  this  latest  period  of  volcanic 
activity  are  preserved  for  our  study  is  not  difficult  to  explain.  There 
is  clear  evidence  that  while  the  eruptions  of  this  period  took  place 
over  a  very  wide  area — being  traceable  from  Yorkshire  on  the  east 
to  Donegal  on  the  west — the  superficial  manifestations  of  the  igneous 
activity  were  usually  small  and  inconspicuous.  Indeed,  it  is  only 
at  a  few  points,  like  Beinn  Hiant  in  Ardnamurchan  and  the  Sgiir 
of  Eigg,  that  the  erupted  material  was  poured  out  in  such  abundance 
as  to  escape  removal  by  subsequent  denudation.  I  long  ago  pointed 
out  that  these  widespread  ejections  must  have  resembled  the  puys 
of  Auvergne,  and  have  consisted  of  cinder-cones  and  lava-currents 
ejected  at  a  number  of  points,  along  lines  of  fissure  which  radiated 
from  the  great  eruptive  centres,  and  were  probabty  developed  after 
the  central  volcanoes  had  become  extinct. 
"While,  however,  most  of  the  volcanic  materials  poured  out  at  the 
surface  have  long  since  been  swept  away  by  denudation,  the 
dykes,  or  filled-up  fissures  beneath  the  volcanic  vents,  remain  as 
witnesses  to  the  number  and  widespread  character  of  these  latest 
eruptions.  In  the  study  of  these  dykes  we  find  abundant  and  very 
interesting  evidence  of  the  process  of  differentiation  of  lavas  ;  nor 
can  it  be  doubted  that  several  kinds  of  lava  were  successively  poured 
out  from  certain  of  the  vents  supplied  from  these  fissures. 
III.  The  Chief  Types  of  Volcanic  Bocks  eepeesented  in 
THESE  LATE  EJECTIONS. 
The  materials  found  in  the  latest  volcanic  djffies  of  the  British 
Islands,  which  traverse  alike  the  Secondary  and  all  older  strata  as 
well  as  the  igneous  rocks  of  the  Hebrides  and  the  Aorth  of  Ireland, 
represent  two  very  distinct  types  of  rock. 
The  first  of  these  has  been  admirably  described  by  Hr.  Teall  in 
his  account  of  the  A orth-of-England  dykes.1  In  the  Cleveland 
dyke  and  similar  eruptive  masses  of  late  date  we  find  a  series  of 
rocks  which  may  be  classed  with  the  augite-andesites.  They  seldom, 
if  ever,  contain  free  olivine,  the  chief  porphvritic  constituents  being 
a  lime-soda  felspar,  which  is  usually  zoned,  and  a  magnesian  augite  ; 
while  their  silica-percentage  ranges  from  52  to  58.  They  sometimes 
contain  but  little  vitreous  matter,  and  in  these  cases  not  unfrequently 
assume  an  ophitic  structure  and  may  be  classed  with  the  dolerites 
(diabases).  When  a  small  quantity  of  glass  is  present,  it  is  often 
seen  to  be  caught  up  between  the  felspar-crystals,  thus  giving  rise 
to  the  structure  styled  by  Bosenbusch  6  intersertal,’  and  entitling  the 
rocks  to  be  called  in  that  petrographer’s  nomenclature  ‘  tholeites.’ 
1  Quart.  Journ.  Geol.  Soc.  vol.  xl.  (1884)  p.  209;  see  also  1  British  Petro¬ 
graphy,’  1888,  pp.  200-207. 
