58  SMITHSONIAN    MISCELLANEOUS    COLLECTIONS  VOL.    96 
solid  matrix  of  permanently  frozen,  rich,  black  soil.  This  is  a  con- 
dition that  would  naturally  result  when  a  kitchen  midden  was  formed 
as  this  one  was — by  the  daily  accumulation  of  rubbish  thrown  just 
outside  the  houses.  In  appearance  such  a  midden  differs  markedly 
from  the  more  familiar  type  found  all  along  the  coasts  as  far  north 
as  the  Aleutian  Islands.  The  principal  point  of  difference  is  that  these 
more  typical  middens  to  the  southward  contain  more  or  less  distinct 
strata  which  are  due  in  large  part  to  the  presence  of  shells,  ashes,  sand, 
and  other  beach  material,  none  of  which  is  present  to  any  appreciable 
extent  in  an  Eskimo  midden  of  the  Miyowagh  type. 
The  first  cuts  at  Miyowagh  were  made  in  the  northwestern  section 
of  the  midden  which  was  slightly  higher  and  had  also  suft"ered  less 
from  the  Eskimos'  digging  than  had  the  southeastern  section.  Fur- 
thermore, it  was  relatively  dry  and  well  drained,  whereas  much  of 
the  southeastern  part  of  the  midden  was  in  a  soggy  condition  and  so 
remained  until  midsummer  when  the  snow  from  the  hillside,  which 
fed  the  small  stream  at  the  eastern  edge  of  the  midden,  had  almost 
disappeared.  As  stated  above,  the  cuts  were  12  feet  square  and  were 
taken  down  in  layers  of  from  2  to  6  inches  in  thickness,  except  for 
thicker  surface  layers  of  sod  or  sterile  layers  of  gravel.  In  the  follow- 
ing pages  descriptions  of  individual  cuts  will  be  reduced  to  a  mini- 
mum. Many  of  the  cuts  yielded  between  200  and  400  specimens,  in 
addition  to  quantities  of  animal  bones,  potsherds,  broken  slate  blades, 
and  other  fragmentary  artifacts.  In  the  brief  descriptions  of  indi- 
vidual cuts  only  those  artifacts  having  definite  chronological  signifi- 
cance will  be  mentioned.  Later,  when  describing  the  dift'erent  types 
of  objects,  significant  distributions  according  to  cut  and  depth  will 
be  noted.  Cuts  1-4,  6,  and  17  (season  of  1930)  were  sunk  in  the 
northwestern  section  of  the  midden,  and  cuts  5,  7-9,  9  a,  9  b,  10,  10  a, 
11-16  in  the  southeastern  section  (text  fig.  10).  In  193 1,  Mr.  Cham- 
bers continued  the  work  on  cut  17,  and  cuts  19,  20,  22,  and  27  were 
sunk  in  the  northwestern  section  of  the  midden,  and  cuts  18,  21,  23-26, 
28,  and  29  in  the  southeastern  section.  Three  houses  were  excavated 
at  Miyowagh  in  1930,  no.  5  in  the  northwestern  section  and  nos.  3 
and  4  in  the  southeastern. 
CUTS    IN    THE    NORTHWESTERN    SECTION 
Cut  I  (pi.  6,  fig.  I,  the  excavation  in  foreground). — Cut  i  was 
5  feet  3  inches  deep;  taken  down  in  10  levels.  Surface  covered  with 
scattered  grass,  no  sod.  The  upper  18  inches  consisted  of  almost  pure 
gravel,  with  sporadic  patches  of  midden  material ;  8  inches  deep  was 
