NO.    I  ARCHEOLOGY    OF    ST.    LAWRENCE    ISLAND — COLLINS  35 
unknown  to  the  Eskimos,  was  completely  buried  beneath  the  sod,  moss, 
and  stones,  and  in  outward  appearance  was  a  normal  section  of  the 
hillside. 
At  these  five  sites  excavations  were  carried  on  from  June  21  until 
October  9,  1930,  when  the  freezing  of  the  ground  finally  brought  the 
work  to  a  close.  During  the  greater  part  of  the  time  our  force  con- 
sisted of  the  writer,  his  assistant,  James  A.  Ford,  and  three  Eskimos. 
Silook,  Maskin,  and  Soonogoruk,  with  a  few  other  Eskimos  occasion- 
ally employed.  From  August  14  to  September  19,  during  my  absence 
on  a  trip  to  the  interior  of  Seward  Peninsula,  Mr.  Ford  was  in  charge 
of  the  work.  In  193 1  Moreau  B.  Chambers  was  intrusted  with  the 
continuation  of  the  work  at  Gambell  and  carried  on  excavations  from 
June  20  to  September  17  with  two  Eskimo  helpers  at  the  sites  of 
Miyowagh,  levoghiyoq,  and  Seklowaghyaget. 
The  method  of  excavation  was  determined  by  the  frozen  condition 
of  the  soil  and  the  time  at  our  disposal.  There  were  four  large  middens 
to  be  investigated  in  addition  to  the  buried  site  on  the  Hillside,  dis- 
covered later ;  with  less  than  4  months  in  which  to  work,  it  clearly 
would  not  have  been  possible,  even  if  we  had  so  desired,  to  excavate 
even  one  of  the  middens  in  its  entirety.  Consequently  we  adopted  the 
method  of  sampling,  sinking  pits  of  uniform  size  at  various  parts  of 
the  middens  in  order  to  obtain  as  representative  a  lot  of  material  as 
possible  from  each.  The  cuts,  except  those  for  the  excavation  of 
houses,  were  all  started  as  pits  12  feet  square  and  in  most  cases  were 
continued  to  the  bottom  as  such.  This  method  was  followed  at  the 
four  middens  which  were  visible  when  we  began  operations,  but  not 
at  the  newly  discovered  Hillside  site  where  conditions  were  different 
and  where  it  was  necessary  first  of  all  to  determine  the  extent  of  the 
accumulation.  The  excavations,  as  always  in  these  latitudes,  were 
handicapped  to  a  considerable  extent  by  the  frozen  ground,  for  the 
soil  remains  frozen  the  year  around,  with  only  a  thin  surface  layer 
thawing  out  during  the  summer.  Artificial  methods  of  thawing  by 
the  use  of  steam  or  cold  water,  such  as  are  employed  in  mining,  are 
not  only  expensive  and  impracticable  but  have  the  added  disadvantage 
of  being  destructive  to  the  more  fragile  objects.  The  most  satisfactory 
method  was  found  to  be  that  of  taking  the  excavations  down  little 
by  little  as  the  frozen  soil  thawed  out  through  exposure  to  the  at- 
mosphere. The  rate  at  which  an  exposed  frozen  surface  would  thaw 
depended  on  the  nature  of  the  material :  a  midden  section  consisting 
of  a  compact  mass  of  soil,  bones,  and  baleen  would  thaw  usually  less 
than  an  inch  a  day,  whereas  sections  in  which  gravel  predominated 
thawed  three  or  four  times  as  fast.    As  a  rule  each  cut  was  worked 
