NO.    I  ARCHEOLOGY    OF    ST.    LAWRENCE    ISLAND — COLLINS  I9I 
as  a  rather  deep  depression  ;  excavation  shinved  about  2  feet  of  g^ravel 
over  the  center  and  about  4  feet  at  the  corners.  Tlie  house  was  not 
completely  excavated,  since  it  belonged  to  the  same  type  as  several 
much  better  preserved  houses  previously  excavated  at  the  eastern  end 
of  St.  Lawrence  Island.  This  type  is  illustrated  in  plate  2,  figure  3, 
and  plate  3.  figure  4.  House  no.  10  diiifered  from  all  of  the  older 
(janibell  houses  in  having  a  wooden  fioor  and  sleeping  platforms,  fea- 
tures also  present  in  the  Punuk  Island  houses.  Our  excavation  un- 
covered only  the  floor  (pi.  64,  fig.  2),  parts  of  the  sleeping  platforms 
on  three  sides,  and  the  entrance,  which  faced  to  the  east.  The  floor 
had  been  built  of  split  logs  neatly  fitted  together,  with  flat  sides  up ; 
it  measured  7  feet  6  inches  north  and  south  by  7  feet  3  inches  east 
and  west.  The  platforms,  which  extended  along  three  sides,  had 
been  built  of  similar  timbers  laid  horizontally  across  timbers  extend- 
ing out  from  the  wall  along  the  floor.  The  rear  wall,  which  was  only 
partially  uncovered,  had  been  constructed  of  upright  timbers,  whale 
jaws,  and  a  few  stones.  The  passage  had  floor  and  walls  of  stones 
like  those  of  houses  no.  8  and  9,  but  no  anteroom. 
The  material  found  in  house  no.  10  was  modern  in  every  respect, 
including  a  large  quantity  of  metal — hoop  iron,  files,  large  knife  blades, 
and  sheets  of  copper.  There  were  also  numerous  fragments  of  modern 
type  lamps,  lamp  rests,  and  cooking  pots,  a  stone  sinker  for  a  tom 
cod  net.  an  ivory  bird  figure,  blunt  arrow  tip  of  ivory,  whetstones, 
and  long,  flat  sledge  shoes  made  of  whale  bone. 
At  Mirrukta,  a  deserted  village  about  3  miles  east  of  Gambell,  ruins 
of  stone-walled  houses  somewhat  dilTerent  from  those  described  above 
were  found.  These  may  be  somewhat  more  recent  than  most  of  the 
stone-walled  structures  at  Gambell,  as  they  are  in  a  better  state  of 
preservation.  They  are  about  18  feet  square,  and  the  walls,  built  en- 
tirely of  stones,  are  mostly  above  ground  (pi.  64,  fig.  5).  We  did 
not  excavate  any  of  them,  so  the  depth  of  the  floor  beneath  the  surface 
is  uncertain,  but  it  was  probably  no  more  than  a  foot.  A  stone  wall 
extending  the  length  of  the  room  formed  an  inner  partition  about  6 
feet  wide,  the  purpose  of  which  is  uncertain  unless  it  served  as  a 
storage  space.  Some  of  the  houses  had  attached  at  one  corner,  on 
the  outside,  a  roughly  circular  stone-walled  cache.  The  roof  con- 
struction could  not  always  be  determined,  but  in  some  cases  two  large 
parallel  whale  jaws  had  formed  the  main  framework  of  the  roof,  which 
had  no  doubt  been  formed  of  walrus  skins.  These  houses  lacked  the 
long  sunken  entrance  passages  such  as  were  attached  to  the  other 
early  forms  of  St.  Lawrence  dwellings ;  the  only  entrance  now  visible 
was  a  low  opening  in  the  east  wall  of  one  of  the  houses  across  which 
