220  PRE-CAMBRTAN    GEOLOGY    OF    NORTH    AMERICA. 
to  have  a  direct  transition  into  the  green  schists  of  the  Keewatin. 
The  green  schist  has  a  nearly  vertical  cleavage.  The  schists  do  not 
always  follow  the  course  of  the  granite  range.  They  are  unconform- 
ably  covered  in  many  places  by  the  quartzite.  The  quartzite  never 
has  a  high  dip.  Near  the  base  it  contains  pebbles  of  quartz  and 
granite,  as  well  as  jasper  and  greenstone.  This  quartzite  is  correlated 
with  the  Pewabic  quartzite  of  Gunflint  Lake,  the  Pokegama  quartzite 
of  the  Mississippi  River,  that  of  Sioux  Falls,  S.  Dak.,  and  that  of 
Baraboo,  Wis.  Conformable  with  the  quartzite  is  the  iron-ore  and 
taconite  horizon.  The  strata  are  siliceous  and  calcareous,  and  are 
banded  with  oxide  of  iron  in  beds  of  variable  length  and  thickness. 
The  ore  is  sometimes  magnetite  and  sometimes  hematite.  To  the 
banded  jaspery  quartzite  associated  with  the  ore  the  term  taconite  is 
applied.  The  greenish  siliceous  slates  or  cherts  constitute  a  transi- 
tion stage  between  the  rocks  of  the  iron  horizon  and  the  black  slates. 
There  is  a  considerable  mixture  of  greenish  material,  apparently  of 
eruptive  origin.  The  greater  part  of  the  rock  is  a  red,  yellow,  black, 
white,  or  green  chert,  sometimes  having  a  thickness  of  200  or  300  feet. 
It  often  has  a  peculiar  brecciated  appearance,  having  been  shattered 
into  angular  fragments  and  recemented  by  the  same  amorphous  silica. 
The  same  fracturing  is  also  visible  in  the  iron  ore.  The  siliceous 
slates  and  cherts  pass  upward  into  a  carbonaceous  argillite  of  great 
thickness,  having  a  dip  varying  from  the  horizontal  to  20°  to  the 
south  or  southwest.  Locally  the  dip  is  as  high  as  45°,  in  which  case 
the  ore  deposits  lie  close  to  the  green  schists.  The  gabbro  flow  is  over 
all  of  the  previous  strata.  The  eifect  of  the  heat  on  the  molten  gab- 
bro was  to  make  the  iron  ore  which  already  existed  in  the  rocks  hard 
and  magnetic.  There  is  good  reason  to  believe  that  the  iron-ore 
deposits  in  their  present  condition  have  been  principally  formed  since 
the  gabbro  overflow.  The  ore  deposits  occur  as  regular  beds,  which 
lie  in  almost  their  original  positions,  usually  having  a  dip  of  less 
than  30°  and  passing  into  the  jaspery  quartzite  or  taconite  in  three 
directions,  and  occasionally  on  all  sides. 
The  theory  of  Irving  as  to  the  origin  of  the  Gogebic  ores  is  par- 
tially adopted.  The  quartzite  is  impervious  to  surface  infiltration. 
The  ore  is  regarded  as  produced  by  chemical  replacement  of  some 
mineral,  chiefly  silica,  by  oxide  of  iron.  As  evidence  of  this,  all  stages 
of  the  process  may  be  seen.  Iron  carbonate  is  found  in  the  Mesabi 
rocks,  but  it  does  not  appear  in  sufficient  quantity  to  permit  the  as- 
sumption that  the  source  of  the  ore  was  originally  a  carbonate.  The 
solvent  for  the  silica  was  probably  carbon  dioxide,  and  its  source  may 
have  been  the  atmosphere,  the  black  slates,  recently  decaying  vegeta- 
tion, or  the  ore  deposits  higher  up  the  hill.  The  silica  removed  from 
the  location  of  the  iron  ores  has  been  added  to  the  grains  of  quartz  in 
the  quartzite,  has  been  deposited  as  chalcedonic  and  flinty  silica,  and 
