626  PRE-CAMBEJAN    GEOLOGY   OF    NORTH    AMERICA. 
netic  oxide  of  iron,  confined  to  the  southern  counties  of  the  Highlands 
and  forming  masses  in  gneiss  and  hornblendic  gneiss  rocks  which 
might  be  called  beds,  but  which  are  thought  to  be  veins.  Their  course 
is  parallel  to  the  layers  of  rock,  but  in  several  instances  after  continu- 
ing with  this  parallelism  for  a  certain  distance  the  ore  crosses  a 
stratum  of  rocks  and  then  resumes  its  parallelism,  and  then  obliquely 
crosses  another,  and  so  on.  Also  in  other  places  where  there  are  great 
beds  of  ore  a  few  small  strips  of  ore  penetrate  the  surrounding  rocks 
as  if  they  have  been  cracked  asunder  and  these  seams  forced  up  from 
the  main  mass  below. 
The  rocks  that  are  most  metamorphosed  are  usually  near  granite, 
syenite,  trap,  quartzose  and  metalliferous  protrusions,  dikes,  and 
veins.  It  is  believed  that  trappean  injections  took  place  as  late  as 
the  time  of  the  Red  Sandstone  of  New  Jersey.  The  granitic,  syenitic, 
and  augitic  rocks  appear  to  belong  to  the  epoch  immediately  preced- 
ing the  slates  and  grits  of  the  Champlain  division,  since  they  have 
altered  the  preexisting  rocks  where  they  come  in  contact  up  to  that 
time,  but  no  traces  of  such  changes  are  found  in  the  more  recent  rocks. 
Another  intrusion  of  granite  is  believed  to  have  preceded  the  Red 
Sandstone  of  Rockland  and  New  Jersey,  being  probably  more  recent 
than  the  rocks  of  the  Catskill  division. 
Cozzens,  jr.,70  in  1843,  divides  the  rocks  of  Long  Island  into  gran- 
ite, syenite,  serpentine,  mica  gneiss,  hornblende  slate,  quartz  rock, 
Primitive  limestone,  and  diluvium.  The  distribution  of  all  is  given. 
At  the  Palisades,  on  the  west  side  of  Hudson  River,  the  section  from 
the  base  up  is  granite,  serpentine  (different  from  that  at  Long 
Island),  sandstone,  greenstone  slate,  and  trap.  The  section  of  Staten 
Island  from  the  base  upward  is  granite,  serpentine,  sandstone,  trap  or 
greenstone,  beds  of  iron  ore,  and  diluvium.  At  Donderberg  the  sec- 
tion is  granite,  gneiss,  talcose  slate,  limestone  (called  Transition  lime- 
stone), and  brick  clay. 
Emmons  (E.),71  in  1846,  gives  a  systematic  treatment  of  the  char- 
acter and  relations  of  the  Taconic  system.  The  Taconic  system  is 
held  to  be  below  the  New  York  system,  because  the  base  of  the  latter 
is  perfectly  schistose,  like  that  of  the  former,  and  because  the  mate- 
rial of  the  New  York  system  is  derived  from  the  Taconic.  Again, 
contacts  between  the  Taconic  system  and  the  Calciferous  sandstone 
and  Hudson  River  shales  show  that  the  former  are  unconformably 
below  the  rocks  of  the  New  York  system.  An  evidence  that  the 
Taconic  system  is  newer  than  the  Primary  rocks  is  the  occurrence  of 
porphyritic  quartz  of  the  Taconic  upon  gneiss.  It  is,  then,  not  to  be 
doubted  that  there  is  a  system  of  rocks  lying  between  the  Hoosac 
Mountain  range  and  Hudson  River  of  an  age  posterior  to  the  gneiss 
and  mica  slate  and  anterior  to  the  New  York  system.  It  consists 
throughout  of  beds  of  sedimentary  matter  in  a  state  of  fine  division 
