NO.    I  ARCHEOLOGY    OF    ST.    LAWRENCE    ISLAND COLLINS  J^, 
of  a  whale  jaw  resting  on  one  small  log  and  overlaid  by  another; 
just  inside  was  a  pile  of  four  stones,  the  function  of  which  is  un- 
known. Continuing  from  these  were  two  more  logs,  one  above  the 
other,  followed  by  two  similar  logs  extending  to  the  south  corner, 
braced  on  the  inside  by  a  wooden  stake.  The  last  mentioned  logs  are 
of  particular  interest  in  that  they  were  laid  directly  over  the  fallen 
timbers  of  another  and  earlier  house  (no.  4),  which  is  at  the  right 
in  plate  8,  figure  2,  in  the  immediate  foreground  in  plate  8,  figure  3, 
and  plate  9,  figure  i,  and  in  the  background  in  plate  9,  figure  2. 
The  central  part  of  the  floor  of  house  no.  3  was  covered  with  large 
flat  stones,  but  along  the  walls  the  floor  was  of  earth  except  for  an 
occasional  stone  slab.  Four  feet  from  the  entrance  was  found  a  small 
pit  or  cache  between  the  floor  stones.  It  was  roughly  triangular  in 
outline,  2  feet  long  by  i  foot  wide,  and  was  covered  by  a  piece  of 
whale  scapula.  It  was  3^  feet  deep  and  the  sides  were  lined  with 
stones  and  whale  and  walrus  bones.  The  contents  were  loose  earth 
and  gravel  with  little  refuse.  Five  artifacts  were  found  in  the  cache : 
a  strip  of  baleen  11^  inches  long  by  i^  inches  wide;  an  ivory  pick; 
bone  snow  knife  (pi.  79,  fig.  5)  ;  bone  awl;  and  a  broken  ivory  box 
handle. 
The  material  found  on  and  above  the  floor  of  house  no.  3  was  note- 
worthy in  two  respects  :  ( i )  the  bone  and  ivory,  particularly  the  lat- 
ter, was  very  dark  in  color,  ranging  from  a  deep  brown  to  almost 
black;  it  was  also  brittle  and  poorly  preserved;  (2)  every  harpoon 
head  and  decorated  object  belonged  to  the  early  Punuk  stage,  not  one 
example  of  Old  Bering  Sea  art  being  found.  This  was  in  striking 
contrast  to  the  adjoining  areas  where,  except  for  a  few  Punuk  pieces 
in  the  upper  levels,  only  Old  Bering  Sea  art  was  found.  The  follow- 
ing harpoon  heads  from  cut  9  are  illustrated:  plate  24,  figure  10 
(36  in.  deep)  and  figure  23  (24  in.)  ;  plate  26,  figure  7  (48  in.)  ; 
plate  28,  figure  11  (12  in.)  and  figure  18  (24  in.).  It  is  evident, 
therefore,  that  house  no.  3  dates  from  the  early  Punuk  period.  It  was 
probably  one  of  the  last  houses  occupied  in  the  SE.  section,  which 
no  doubt  accounts  for  the  fact  that  before  excavation,  it  appeared  as 
the  deepest  and  most  prominent  pit  of  any  on  the  midden  proper. 
House  no.  4. — As  stated  above,  an  interesting  case  of  super^xisition 
was  found  at  the  back  of  house  no.  3,  the  south  corner  of  this  house 
having  been  built  over  the  fallen  timbers  of  an  earlier  house  (no.  4). 
The  latter  was  a  smaller,  square  structure,  9  feet  6  inches  E.-W.  by 
9  feet  7  inches  N.-S.  (text  fig.  11).  It  had  a  stone  floor  which  was 
I  foot  lower  than  that  of  the  larger  house  and  a  long  curving  entrance 
passage  which  faced  toward  the  west.  Plate  8,  figures  2  and  3,  and 
plate  9,  figures  i  and  2,  show  the  relationship  of  the  two  houses. 
