GOLD-BEARING    RIVER    SANDS    OF    N.   E.    WASHINGTON.  59 
of  more  recent  origin.  Near  the  west  end  of  the  reservation  there 
are  many  erratic  bowlders  and  other  indications  of  a  glacier  which 
moved  down  the  Okanogan  Valley  from  the  north  and  spread  out  over 
the  big-bend  plateau  to  the  brink  of  the  Grand  Coulee."  The  valley 
of  the  Okanogan  and  the  plateau  south  of  the  Columbia  at  this  point 
are  said  to  be  strewn  with  glacial  drift.  Small  areas  of  bowlders, 
probably  the  terminal  moraines  of  valley  glaciers,  are  found  in  the 
Sanpoil  Valley  about  50  miles  north  of  the  Columbia  and  in  the 
Columbia  Valley  a  few  miles  north  of  the  mouth  of  the  Spokane.6  In 
the  latter  locality  the  moraine  appears  to  be  older  than  the  terraces 
by  which  it  is  apparently  overlain. 
GEOLOGIC    HISTORY. 
The  geologic  history  of  the  development  of  these  topographic  forms 
is  interesting  and  may  be  briefly  summarized  here.  Previous  to  the 
middle  of  the  Tertiary  period  the  drainage  of  northern  Washington 
was  developed.  Columbia,  Sanpoil,  and  Okanogan  rivers  flowed 
southward  across  the  big-bend  country,  converging  at  some  point 
now  buried  beneath  the  basalts/'  Then  followed  a  period  of  volcanic 
disturbance;  the  earth's  crust  south  of  the  Columbia  was  fissured  and 
great  volumes  of  lava  poured  out  and  spread  in  sheets  over  the  sur- 
face, burying  the  old  surface  beneath  thousands  of  feet  of  new  rocks. 
The  pre-Miocene  valleys  were  filled,  and  the  Columbia  was  diverted 
to  a  new  channel  around  the  northern  edge  of  the  lava  field,  inter- 
cepting the  valleys  of  the  Sanpoil  and  Okanogan  as  it  does  now. 
After  the  volcanic  period  the  whole  region  was  probably  elevated  and 
these  rivers  cut  gorges  nearly  if  not  quite  to  the  present  depth.  Later 
the  Okanogan  glacier  completely  dammed  Columbia  River,  forming 
in  its  upper  valley  a  lake  that  extended  for  many  miles.  The  waters 
of  this  lake  spilled  southward  along  the  east  side  of  the  Okanogan 
glacier,  cutting  the  500-foot  gorge  known  as  the  Grand  Coulee/7  whose 
bed  at  its  upper  end  is  cut  into  the  crystalline  rocks  below  the  basalts. 
It  is  estimated  to  be  from  1,000  to  1 ,500  feet  above  the  river  and  indi- 
cates approximately  the  depth  of  water  before  the  lake  disappeared. 
While  the  glacier  remained  the  lake  thus  formed  was  nearly  filled 
with  sediments,  the  remnants  of  which  form  the  higher  terraces  and 
benches  above  described.  The  upper  terrace  of  the  Columbia  is 
reported  to  be  continuous  with  a  stream  terrace  in  the  Grand  Coulee. 
When  the  ice  barrier  disappeared,  the  river  resumed  its  former  course, 
o  Russell,  I.  C,  Geological  reconnaissance  in  central  Washington:  Bull.  U.  S.  Geol.  Survey  No.  108, 
1893,  pp.  87-89.  Smith,  G.  O.,  and  Calkins,  F.  C,  Bull.  U.  S.  Geol.  Survey  No.  235,  1906,  pp.  35-41.  Salis- 
bury, R.  D..  Am.  Geologist,  vol.  9,  1902,  pp.  212-213. 
''Salisbury,  R.  I).,  Glacial  work  in  western  mountains  in  1901:  Jour.  Geol.,  vol.  9,  1901,  pp.  718-731. 
r  Willis,  Bailey,  Changes  in  river  courses  in  Washington  Territory  due  to  glaeiation:  Bull.  TJ.  S.  Geol. 
Survey  No.  40,  1887,  pp.  7-8. 
d  Russeil,  1.  C,  Bull.  U.  S.  Geol.  Survey  No.  108  1893,  pp.  90-92. 
