638  PRE-CAMBRTAN    GEOLOGY    OF    NORTH   AMERICA. 
supposed  dikes  of  granite  are  strata  conformable  to  the  white  lime- 
stone, as  are  the  iron  and  zinc  ore  beds  contained  in  it,  all  geologically 
older  than  the  blue  limestone  with  the  quartzites  and  slates  composing 
the  Lower  Silurian  system.  The  conclusion  is  now  reached  that  the 
unstratified  rock  masses  underlie  the  bedded  crystalline  rocks,  al- 
though the  line  of  separation  is  but  poorly  defined,  as  the  stratified 
rocks  of  the  same  mineral  composition  commonly  occur  on  the  sides  of 
the  massive  area,  with  an  apparent  gradual  passage  between  the  two, 
and  at  no  point  was  any  actual  unconformity  found,  although  at  some 
places  abrupt  changes  in  the  lamination  were  observed  within  short 
distances.  This  leads  to  the  conclusion  that  the  massive  beds  are 
so  only  because  stratification  has  been  wholly  destroyed  through 
greater  metamorphism.  The  schistose  series  commonly  have  a  steeper 
dip  along  their  southern  margins  than  along  their  opposite  sides; 
thus  the  axial  planes  of  the  folds  are  often  inclined  toward  the  south- 
east. 
While  the  Potsdam  and  Paleozoic  rocks  are  unconformable  upon 
the  Archean,  the  newer  rocks  are  tilted  in  such  a  way  as  to  show  that- 
folding  has  occurred  since  they  were  deposited.  At  only  a  few  places 
are  actual  junctions  found,  the  two  more  important  being  in  Owens 
Island,  in  Sussex  County,  and  at  Franklin  Furnace.  At  several  lo- 
calities the  relations  are  perplexing,  for  the  quartzites  and  conglom- 
erates are  so  heavily  feldspathic  that  near  the  junction  they  appear 
to  grade  gradually  into  the  older  rocks,  fragments  and  masses  of 
which  are  included  in  them.  Along  the  southeastern  margin  of  the 
Highlands  the  Silurian  is  crystalline,  including  crystalline  limestones 
and  hydromica  slates,  and  here  the  unconformity  is  much  less  pro- 
nounced, no  satisfactory  contacts  being  known  in  New  Jersey.  At 
Pompton  the  slate  ledges  have  nearly  the  same  dip  and  strike  as  the 
nearest  Archean  outcrops.  At  Peekskill  Hollow  and  Annsville  Cove, 
in  New  York,  the  slates  and  quartzites  and  crystalline  rocks  appear 
to  be  directly  conformable,  the  strata  having  been  subjected  to  an 
overturn  that  caused  the  quartzite  to  dip  under  the  older  rocks,  and 
it  is  difficult  to  say  where  the  line  of  separation  is. 
Martin,104  in  1886,  states  that  the  Tidewater  gneiss  has  mineralog- 
ical  characters  which  distinguish  it  strongly  from  the  gneisses  of  the 
northern  Laurentian  and  from  the  Highlands.  In  particular  the 
abundance  of  subsilicates  and  of  hydrous  silicates  is  to  be  noted. 
Britton,105  in  1887,  divides  the  Archean  rocks  into  a  massive  group, 
an  iron-bearing  group,  and  a  gneissic  and  schistose  group,  which  is 
also  believed  to  be  the  order  of  superposition,  although  there  is  a 
gradual  change  from  one  sedimentary  rock  into  the  other.  While  the 
massive  rocks  are  but  faintly  laminated,  there  is  no  evidence  adduci- 
ble  in  favor  of  an  igneous  origin  for  them,  but  all  indications  point  to 
their  deposition  as  sediments  of  one  kind  or  another,  and  to  the  more 
