Yol.  49.  ]  THE  LLEYX,  AND  ASSOCIATED  YOLCAHIC  EOCKS.  157 
low  temperature.1  They  would  often  occur  where  some  freedom  of 
motion  is  possible,  as  near  the  exterior  of  the  rock-mass,  at  the 
boundary  of  contraction-spheroids,  or  along  the  lines  of  flow  in  a 
glassy  magma. 
In  spheroids,  which  we  may  consider  as  a  typical  case,  the  order 
of  changes  might  be  something  like  the  following  : — The  exterior 
forming  a  thin  film,  the  rate  of  cooling  in  the  next  shell  is  so 
checked  that  crystallites  begin  to  develop.  In  the  viscid  crust, 
iron  oxide  segregates  at  places  constituting  centres  of  growth. 
Badial  and  concentric  cracking  takes  place,  and  further  cooling 
causes  the  glass  to  ‘  set  ’  along  the  fissures.  In  the  ferruginous 
aggregates  within  the  crust,  the  constituents  group  themselves,  and 
a  radial  arrangement  dominates  the  growth,  thus  giving  rise  to 
spherulites.  The  interior  of  the  spheroid  is  still  in  a  plastic  con¬ 
dition,2  and  as  soon  as  the  crust  is  sufficiently  solid,  the  fall  of 
temperature  in  the  part  within  takes  place  doubtless  at  a  fairly 
uniform  and  somewhat  slower  rate,  so  that  a  confused  intercrystalli¬ 
zation  of  its  elements  results.  At  certain  parts  they  are  connected 
together  in  radial  groups,  but  these  are  not  definitely  bounded,  thus 
they  do  not  weather  out,  and  the  rock  where  they  occur  is  not  a 
variolite. 
The  same  succession  of  conditions  would  be  found  towards  the 
exterior  of  a  rock-mass,  so  that  a  variolite  is  limited  sometimes  to 
that  part.  A  somewhat  analogous  case  might  occur  in  a  moving 
lava-stream,  where  semi-plastic  fragments  are  embedded  in  a  more 
fluid  part.  The  movement  would  tend  to  prevent  the  proper  forma¬ 
tion  of  spherulites,  but  the  curious  aggregations  along  the  exterior 
of  included  pieces  might  be  an  incipient  development.3 
Some  of  the  Lleyn  variolite  is  thus  a  development  of  a  lava ;  and 
where  the  magma  was  not  an  actual  flow,  it  seems  very  probable 
that  it  cooled  at  no  great  depth  below  the  surface,  but  was  formed 
towards  the  exterior  of  an  intrusive  mass.4 
Certain  of  the  spherulites  (as  for  example  that  from  Deunant) 
are  not  unlike  the  variolite  of  Brian  con,  a  specimen  of  which 
was  kindly  lent  me  for  comparison  by  Prof.  Bonney  ;  but  in  the 
Durance  rock  the  crystallization  is  more  definite,  and  a  chloritic 
constituent  has  developed,  which  seems  to  have  been  cleared  out 
of  the  spherulite,  so  that  this  is  paler,  but  is  surrounded  by  a 
darker  rim. 
1  Quart.  Journ.  Geol.  Soc.  vol.  xlvii.  (1891)  p.  105.  Compare  also  experi¬ 
ments  of  M.  Claudet,  referred  to  ibid.  vol.  xli.  (1885)  Proc.  p.  90,  note;  Cole 
and  Gregory,  ibid.  vol.  xlvi.  (1890)  p.  326  ;  J.  W.  Gregory,  ibid.  vol.  xlvii.  (1891) 
p.  61. 
2  This  agrees  with  Mr.  J.  W.  Gregory’s  suggestion,  ibid.  vol.  xlvii.  (1891) 
p.  60.  See  also  Cole  and  Gregory,  ibid.  vol.  xlvi.  (1890)  p.  316,  and  Delesse, 
‘  Metamorphism  e  des  Roches,’  Paris,  1858,  pp.  371-374. 
3  Some  structures  at  Porth  Oer  are  probably  due  to  similar  conditions. 
The  development  is  shown  clearly  in  a  specimen  from  Boulay  Bay,  which  I 
received  from  Prof.  Bonney  for  comparison  with  nodules  obtained  near 
Pwllheli.  See  Quart.  Journ.  Geol.  Soc.  vol.  xlv.  (1889)  p.  264. 
4  See  Prof.  Cole’s  suggestion  as  to  the  Annalong  dyke,  Sci.  Proc.  Roy.  Dub. 
Soc.  vol.  vii.  (1892)  p.  513. 
