Yol.  49.]  PEOF.  J.  W.  JUDD  ON  COMPOSITE  DYKES  IN  AEEA.N.  537 
(B)  Composite  dykes  in  which  there  has  been  injection  of  different 
materials  into  the  same  fissure. — In  these  the  later  injected  rock, 
though  it  may  contain  minerals  and  fragments  derived  from 
the  older  injection,  exhibits  no  trace  of  gradation  into  it,  but 
can  always  be  recognized  as  another  type  of  material,  distinct 
in  chemical  composition  or  mineralogical  constitution.  In  such 
cases  we  may  infer  that  differentiation  must  have  taken  place  in 
the  magma  deep  down  in  the  earth's  crust,  and  far  below  the 
fissure  which  is  now  filled  with  the  two  products  resulting 
from  the  differentiation.  While,  however,  it  is  probable  that 
such  composite  dykes  are  usually  filled  with  materials  coming 
from  the  same  subterranean  reservoir,  it  is  possible  that,  in 
certain  exceptional  cases,  a  dyke  may  be  reopened  in  such  a 
way  as  to  allow  of  molten  material  from  a  source  totcdly  dif¬ 
ferent  from  that  which  supplied  the  earlier  injection  finding  its 
way  into  the  fissure.  In  such  exceptional  cases  the  composite 
character  of  the  dyke  must  be  regarded  as  an  accident. 
In  many  instances,  the  younger  injection  is  found  occupying 
the  centre  of  the  dyke ;  for,  owing  to  the  cooling  of  the  mass 
from  the  sides,  a  plane  of  weakness  (often  very  manifest  in 
columnar  dykes)  may  originate  along  the  middle  of  the  dyke ; 
in  this  plane,  indeed,  contraction  due  to  crystallization  not 
unfrequently  leads  to  the  formation  of  numerous  cavities. 
But  there  are  other  instances  where  the  junctions  between 
the  dykes  and  the  enclosing  rock  constitute  planes  of  weakness, 
along  which  the  later  ejected  material  has' forced  its  way* 
In  such  cases,  the  older  rock  occupies  the  centre  of  the  dyke. 
While  the  two  classes  of  composite  dykes  to  which  I  have  just 
referred  are  usually  very  distinct  and  clearly  recognizable,  the  one 
from  the  other,  there  are  certain  examples  which  seem  to  constitute 
a  link  between  them.  These  are  the  composite  dykes  in  which  the 
dyke-rock  appears  to  be  intersected  by  veins,  of  the  nature  of  the 
‘  contemporaneous  ’  or  ‘  segregation-veins  ’  so  common  in  plutonic 
rocks.  It  is  difficult  in  such  cases  to  be  certain  whether  the 
differentiation  has  taken  place  in  the  dykes  themselves,*  or  in  the 
reservoirs  beneath,  from  which  they  have  been  supplied. 
The  question  of  the  order  of  succession  of  different  lavas  issuing 
from  a  volcanic  vent  is  one  to  which  the  attention  of  geologists  has 
long  been  directed.  Quite  recently,  however,  interest  in  the  subject 
has  been  revived,  owing  to  certain  theoretical  suggestions  having 
been  put  forward  to  account  for  the  observed  facts.  It  is  obvious 
that  the  study  of  the  succession  of  igneous  rocks  at  any  centre  of 
eruption  may  throw  light  upon  the  processes  of  differentiation, 
which  appear  to  have  been  carried  on  within  many  molten  rock- 
masses.1  In  connexion  with  this  important  problem,  the  study  of 
1  For  a  summary  of  the  wort  already  done  in  connexion  with  this  subject 
Mr.  Iddings’s  memoir  ‘  On  the  Origin  of  Igneous  Bocks,’  Bull.  Phil.  Soc. 
AYashington,  vol.  xii.  (1892)  pp.  89-214,  may  be  consulted. 
