372  SMITHSONIAN    MISCELLANEOUS    COLLECTIONS  VOL.    96 
stages  of  culture  in  northern  Alaska  rather  than  with  the  Birnirk 
or  Old  Bering  Sea  phases,  we  have  still  to  account  for  its  origin. 
In  seeking  this  origin  in  northern  Alaska  Mathiassen  has  arrived 
at  what  seems  to  be  the  only  possible  solution;  it  remains  to  be  de- 
termined from  just  what  region  and  what  culture  stage.  Our  knowl- 
edge of  archeological  conditions  in  this  region  is  insufficient  to  permit 
of  a  definite  answer,  but  according  to  present  indications,  when  the 
Thule  left  the  western  regions  it  must  have  been  at  a  stage  somewhat 
later  than  that  of  the  Birnirk  culture.  The  fairly  large  number  of 
basic  common  elements  in  the  two  cultures  attests  their  general  rela- 
tionship, but  there  is  still  a  wide  gap  between  the  oldest  of  the  Eastern 
Thule  sites — Naujan  and  Malerualik — and  the  two  old  Barrow  sites. 
This  suggests  that  somewhere  in  the  intervening  area  there  are  to  be 
found  older  Thule  sites  than  those  excavated  by  Mathiassen,  sites  that 
will  show  a  more  pronounced  resemblance  to  Birnirk,  such  as  possibly 
in  a  more  general  use  of  earthenware  pottery  and  of  harpoon  heads 
closer  to  the  Birnirk  types.  At  the  same  time  one  would  expect  that 
these  intermediate  sites  would  not  reveal  in  fully  developed  form  those 
Thule  features  which  were  later  introduced  into  Alaska. 
The  Birnirk  culture  itself  must  presumably  be  regarded  as  a  direct 
outgrowth  of  the  Old  Bering  Sea  culture,  a  somewhat  sjiecialized 
phase  that  became  established  at  the  eastern  periphery.  A,  basic  rela- 
tionship with  the  Old  Bering  Sea  culture  is  shown  by  the  considerable 
number  of  implement  types  common  to  the  two  stages  (many  of  those 
which  are  mentioned  by  Mason,  p.  386),  by  the  sporadic  occurrence 
of  Old  Bering  Sea  art,  and  by  the  harpoon  heads,  which,  like  some 
of  the  earliest  Old  Bering  Sea  types,  have  side  blades  and  prominent 
asymmetrical  divided  spurs.  However,  the  particular  forms  of  these 
Birnirk  heads  are  closer  to  some  of  the  early  Punuk  than  to  the  Old 
Bering  Sea  types.  It  would  appear  that  they  represent  a  special  line 
of  development  from  the  general  Old  Bering  vSea  pattern.  According 
to  the  list  of  features  given  by  Mason  (1930,  p.  386)  there  are  several 
others  which  denote  a  closer  relationship  with  the  Puiuik  than  the 
Old  Bering  Sea :  bola  weights,  sealing  scratchers,  ivory  pendants, 
wrist  guard. 
One  of  the  most  important  problems  of  Arctic  archeology  is  thai 
of  the  origin  and  relationships  of  the  Dorset  culture.  Originally  de- 
scribed by  Jenness  (1925)  from  Cape  Dorset  and  Coats  Island  in 
the  northern  part  of  Hudson  Bay,  it  is  now  known  to  have  extended 
from  north  Greenland  to  Newfoundland  (Jenness,  1933,  pp.  390-395)- 
In  his  last  publication  Jenness  has  shown  conclusively  that  the  Dorset 
