vanhise.]  LAKE    SUPERIOR    REGION.  81 
ments  of  the  breccia  are  but  slightly  rounded  and  worn  on  their  edges, 
having  in  this  respect  the  appearance  of  a  friction  conglomerate.  The 
blocks  are  cemented  together  by  a  hard  ferruginous  paste. 
The  granites  belong  to  two  different  epochs:  That  of  the  northwest- 
ern coast  and  the  vicinity  of  Pigeon  river  was  elevated  before  the  Azoic 
period,  since  the  upper  portion  of  the  slates  repose  horizontally  around 
it ;  while  that  of  the  northeastern  coast,  and  that  which  forms  the  axis 
of  the  river  systems  of  the  two  lakes,  was  elevated  after  the  termination 
of  the  Azoic  period  and  before  the  dawn  of  the  Silurian,  since  the  gran- 
ite has  disturbed  the  upper  beds  of  slate,  while  the  lower  beds  of  the 
Potsdam  rest  undisturbed  around  it. 
The  masses  of  iron  ore  and  jasper  have  none  of  the  characteristics  of 
vein  deposits.  They  are  intercalated  among  the  metamorphosed  sedi- 
mentaries  and  have  an  intimate  association  with  the  trappean,  porpky- 
ritic,  and  serpentine  rocks.  If  the  trappean  rocks  were  an  invariable 
accompaniment,  the  ores  would  with  little  hesitancy  all  be  assigned  to 
a  purely  eruptive  origin;  but  when  they  are  found  in  the  form  of  beds, 
in  clearly  inetamorphic  strata,  having  a  common  bearing  and  inclina- 
tion, they  are  regarded  as  having  been  derived  from  the  destruction  of 
previously  formed  igneous  masses,  and  their  present  association  as  having 
resulted  from  aqueous  deposition.  The  Azoic  period  having  been  one 
of  long  continued  and  violent  mechanical  action,  there  is  no  reason  to 
doubt  that  many  of  the  strata  of  which  it  is  composed  may  have  been 
derived  from  the  ruins  of  previously  formed  rocks  of  the  same  age,  both 
sedimentary  and  igneous,  as  is  shown  by  the  case  of  the  knob  of  con- 
glomerate already  mentioned.  The  minute  banding  of  the  ore  and 
(jasper  can  hardly  be  explained  by  any  other  than  the  action  of  segre- 
gating forces  in  an  igneous  rock.  The  authors  are  then  disposed  to 
jregard  the  specular  and  magnetic  oxide  of  iron  as  a  purely  igneous 
product,  in  some  instances  poured  out,  in  others  sublimed,  from  the 
interior  of  the  earth.  Where  the  ores  are  in  a  state  of  purity,  in  the 
form  of  irregular  masses  in  preexisting  depressions,  or  where  the  in- 
cumbent strata  are  metamorphosed  or  traversed  by  the  dikes  of  ferru- 
ginous matter,  they  are  without  doubt  eruptive.  Where  impregnating 
metamorphic  j)roducts,  such  as  jasper,  hornstone,  or  chert,  quartz, 
phlorite,  and  talc  slate,  not  only  between  the  lamime  but  intimately 
Incorporated  with  the  mass,  giving  it  a  banded  structure,  they  are 
•egarded  as  the  results  of  sublimation.  The  supposition  that  the  ore 
nay  be  a  secondary  product  resulting  from  the  decomposition  of  a 
ayrite,  or  the  metamorphism  of  bog  iron,  is  inadequate  to  account  for 
3he  accumulation  of  such  mountain  masses  and  to  explain  the  relations 
;o  the  associated  rocks. 
The  bed  of  lake  Superior,  embracing  an  area  of  about  32,000  square 
niles,  is  occupied  almost  exclusively  by  the  Potsdam  sandstone     The 
andstone  in  the  vicinity  of  the  trappean  rocks  attains  the  enormous 
hickness  of  5,000  feet,  and  often  consists  of  conglomerates  composed 
Bull.  86 6 
