342  SMITHSONIAN    MISCELLANEOUS    COLLECTIONS  VOL.    96 
from  the  upper  levels  of  Seklowaghyaget,  and  from  houses  8,  9,  and 
10  we  obtained  quantities  of  fragmentary  pottery  of  this  kind — pieces 
of  flat-bottomed  lamps,  oval  to  rectangular  in  shape,  with  two  high 
parallel  ridges  extending  along  the  inside,  like  the  modern  specimen 
shown  in  plate  84,  figure  4 ;  and  of  flat-bottomed,  thin-walled  cooking 
pots,  rectangular  in  shape  with  rounded  corners  and  with  suspension 
lugs  on  the  outside  or  holes  through  the  rim  (pi.  84,  fig.  5).  But  the 
far  more  extensive  excavations  at  levoghiyoq,  Miyowagh,  and  the 
Hillside  site  revealed  no  trace  of  this  kind  of  pottery ;  the  sherds  indi- 
cated that  there  had  been  no  lamps  with  ridges,  no  pots  with  suspension 
holes  or  lugs.  Instead  the  lamps  were  shallow,  conical,  and  thick- 
wallcd,  and  the  cooking  pots  deep  and  circular  with  rounded  or  flat 
bottoms  and  thin  walls.  Although  no  complete  vessels  were  found,  the 
many  hundreds  of  sherds  indicate  that  the  forms  of  both  lamps  and 
cooking  pots  remained  constant  throughout  the  Old  Bering  Sea  period 
and  the  greater  part  of  the  Punuk. 
Unlike  the  modern  St.  Lawrence  lamp,  which  is  peculiar  to  the 
island,  the  older  rounded  form  seems  to  have  been  rather  widely 
distributed  in  northern  Alaska.  I  found  a  number  of  them  when  ex- 
cavating old  burials  at  Kowieruk,  in  the  western  part  of  Seward 
Peninsula  and  also  at  Wales ;  one  of  the  Kowieruk  lamps  is  illustrated 
in  plate  53,  figure  2.  These  lamps  do  not  conform  exactly  to  any 
modern  type,  though  the  rounded  shape  indicates  a  general  relationship 
to  the  saucer-shaped  lamps  of  the  Yukon-Kuskokwim  region.  Mason 
states  that  round,  flat  pottery  lamps  and  round  earthenware  pots  were 
excavated  by  Van  Valin  at  Barrow  (1930,  p.  386).  Stefansson  also 
found  pottery  fragments  to  be  abundant  at  old  ruins  in  the  Horton 
River  district  between  the  Mackenzie  and  Coronation  Gulf  (1913, 
pp.  327-329;  1914,  pp.  312-314,  332,  348),  and  Jenness  found  frag- 
ments in  the  Coronation  Gulf  region  (1923,  p.  541).  Soapstone  ves- 
sels seem  to  have  been  absent  at  all  of  the  old  Alaskan  sites  (Stefans- 
son. 1914.  p.  394;  Jenness.  1928  a,  pp.  74,  75;  Mason,  1930,  p.  386; 
Mathiassen,  1930  a,  p.  53;  Collins,  1934,  p.  311,  and  1935,  p.  463). 
The  archeological  evidence  clearly  indicates,  therefore,  that  pottery 
lamps  (and  also  cooking  pots),  instead  of  being  restricted  to  the 
Yukon-Kuskokwim  region  and  St.  Lawrence  Island,  were  until  quite 
recently  uniformly  distributed  throughout  northern  Alaska  as  well. 
In  fact  I  do  not  know  of  an  abandoned  Eskimo  site  from  Norton 
Sound  to  Point  Barrow  at  which  pottery  fragments  cannot  be  found 
in  large  numbers.  According  to  all  indications,  pottery  persisted  in 
most  places  up  to  a  few  generations  ago.  The  old  lamps  did  not  dis- 
appear immediately  upon  the  importation  of  soapstone  lamps  from  the 
