van  hise.]  LAKE    SUPERIOR    REGION.  131 
gabbro.  The  gabbro  is  sometimes  cut  by  veins  or  dikes  of  syenite, 
which  indicates  that  the  latter  is  of  later  age,  although  the  syenite  is 
generally  below  the  gabbro. 
Winchell  (Alex.),181  in  1889,  maintains  that  the  Saganaga  and 
West  Sea  Gull  granite  conglomerate  before  described  is  produced 
from  a  fragmental  rock  by  selective  metamorphism,  the  completely 
crystalline  gneissoid  rocks  retaining  rounded  fragments  which  are 
residual  clastic  material.  The  conglomerate  of  Wonder  island  is  not 
one  consisting  originally  of  a  mass  of  pebbles  over  which  a  fluid 
magma  has  been  poured,  for  the  pebbles  are  not  in  contact;  they  could 
not  have  lain  where  they  are  before  the  gneissic  magma  existed.  The 
gneissic  magma  was  present,  and  it  was  this  which  supported  the  peb- 
bles and  prevented  their  contact.  It  is,  then,  contemporaneous  with  the 
pebbles.  The  magma  must  have  been  plastic,  but  it  was  low  tempera- 
ture igneo- aqueous  plasticity. 
Winchell  (N.  H.),182  in  1889,  in  a  general  discussion  of  the  origin  of 
the  eruptive  rocks,  maintains  that  there  are  four  epochs  of  basic  erup- 
tion in  Minnesota:  first,  the  Vermilion  group;  second,  those  succeeding 
the  graywackes;  third,  those  succeeding  the  Animikie;  and  fourth, 
those  of  the  Cupriferous  formation. 
Meads,183  in  1889,  describes  the  Stillwater,  Minnesota,  deep  well. 
The  well,  after  passing  through  about  700  feet  of  St.  Croix  and  Potsdam 
sandstone,  passes  into  rocks  which  are  in  every  respect  identical  with 
those  of  Keweenaw  point ;  hence  the  Keweenaw  rocks  are  below  the 
light  colored  sandstones  of  the  northwest.  For  the  first  1,500  feet  these 
are  brown  shales  and  brown  feldspathic  sandstones,  and  these  gradu- 
ally assume  the  characters  of  a  volcanic  detrital  tuff — amygdaloid — 
and  finally  at  a  depth  of  3,300  feet  unmistakable  beds  of  trap  were  en- 
countered alternating  with  sandstone  beds.  At  this  depth  grains  of 
native  copper  were  seen  in  the  drillings. 
Winchell  (N.  H.  and  H.  V.),184  in  1889,  maintain  that  the  iron  ores 
of  the  Keewatin  of  Minnesota  are  not  derived  from  a  carbonate,  but  are 
probably  a  direct  chemical  precipitate;  for  there  is  no  evidence  of  the 
existence  of  carbonate  of  iron  at  any  time,  and  the  nature  of  the  coun- 
try rock  is  rach  as  to  imply  that  no  carbonates  in  amounts  required 
could  have  b%n  deposited  at  the  time  the  rocks  were  formed. 
Winchell  (Alex.),185  in  1890,  repeats  his  general  conclusions  as  to 
the  stratigraphy  in  northeastern  Minnesota  and  gives  in  descending 
order  a  succession,  as  follows  : 
V.  The  Uncrystalline  schists  (Animikie,  Huroniau). 
IV.  The  Semi-crystalline  schists  (Keewatin). 
III.  The  Crystalline  schists  (Vermilion). 
II.  The  Gneissoid  rocks. 
I.  The  Granitoid  rocks  (Laurentian). 
Winchell  (N.  H.  and  H.  V.),186  in  1890,  state  that  the  iron  ores  of 
Minnesota  are  at  five  different  geological  horizons,  in  descending  order, 
