60  CONTRIBUTIONS    TO    ECONOMIC    GEOLOGY,  1906,  PART    I. 
and  its  subsequent  work  has  been  the  removal  of  these  sediments  from 
its  valley.  New  and  lower  flood  plains  have  been  formed  at  interval! 
during  this  restoration  of  the  old  channel,  a  fact  which  accounts  to 
some  extent  for  the  irregularity  of  the  lower  benches. 
All  the  terraces  within  300  feet  of  the  river  level  were  formed  since 
the  disappearance  of  the  greater  part  of  the  barrier  and  under  nearly 
normal  river  conditions. 
The  upper  terraces  consist  of  only  imperfectly  stratified  sands  and 
silts.  Coarse  gravels  are  comparatively  rare  and  no  false  bedding  has 
been  observed.  The  lower  benches,  on  the  other  hand,  are  charac- 
terized by  much  false  bedding.  In  many  places  they  are  compose! 
of  coarse  gravels  and  sands  and  have  all  the  characteristics  of  river 
deposits. 
RELATION  OF  GOLD  DEPOSITS  TO  TERRACES. 
Where  observed  the  placer  gold  along  the  Columbia  is  confined  to 
the  lower  benches  and  river  bars,  a  condition  which  may  reasonably 
be  expected  if  the  sediments  of  the  upper  terraces  are  lake  deposits, 
and  those  of  the  lower  terraces  have  been  worked  oxer  and  concen- 
trated by  the  river.  Moreover,  the  later-  benches  and  the  modern 
river  bars  are  progressively  richer  in  gold,  since  they  are  the  product 
of  repeated  concentrations  of  the  various  upper  terraces  which  have 
fallen  into  the  river  and  been  washed  away. 
The  terraces  on  which  deposits  of  placer  gold  occur  are  all  within 
200  feet  of  t'ne  level  of  the  river.  They  are  in  the  main  covered  with 
a  sandy  soil  from  a  few  inches  to  12  feet  or  more  t  hick,  containing  some 
disseminated  gold,  below  which  is  a  pay  streak  from  1  inch  to  4  feet 
thick,  consist  ing  of  sandy  clay  and  gravel  rest  ing  on  a  t  bin  bed  of  claw 
Below  this  pay  streak  there  is  generally  barren  sand,  gravel,  or  in  some 
places  clay  to  the  level  of  I  he  river.  It  is  reported  that  a  second  or  a 
third  lower  pay  streak  occurs  at  a  few  places,  but  no  attempts  have 
been  made  to  mine  such  deposits,  and  our  investigations  indicate  that 
they  invariably  contain  less  gold  than  the  upper  placers.  The  pay 
streak  is  as  a  rule  stained  by  oxide  of  iron  and  easily  distinguished 
from  the  overlying  slits  by  its  color.  The  gold  tenor  varies  from  a 
fraction  of  a  cent  to  a  possible  maximum  of  $1.50  per  ton,  the  average 
being  not  more  than  I  or  5  cents  per  cubic  yard.  These  gold  deposits 
were  tested  by  carefully  panning  samples  taken  from  prospect  holes 
and  such  natural  exposures  as  cut  banks  of  streams  and  upturned 
tree  roots.  The  colors  of  gold  in  each  pan  were  counted,  the  num- 
ber recorded,  and  specimens  from  the  various  localities  weighed  on  an 
assayer's  balance  to  determine  their  average  value.  To  check  these 
results  :W  samples  of  pay  dirt  were  collected  and  sent  to  the  Unite! 
Slates  Geological  Survey  concentrating  pavilion  at  Portland,  OregJ 
where  they  were  tested  by  fire  assay.      Jn  making  the  assays  about 
