NO.    I  ARCHEOLOGY   OF    ST.    LAWRENCE    ISLAND — COLLINS  289 
culture,  which  is  manifested  in  a  number  of  important  culture  traits, 
and  which  extends  even  to  linguistics  and  physical  characteristics. 
As  the  writer  has  previously  suggested,  this  is  a  condition  which 
might  be  explained  on  the  assumption  that  there  had  been  a  return 
migration  of  Thule  peoples  to  northern  Alaska  within  the  past  few 
centuries,  subsequent  to  the  original  eastward  spread  of  the  Thule 
culture  (Collins,  1935,  pp.  463,  464).  We  will  return  to  a  considera- 
tion of  this  problem  later. 
If  the  art  of  southwestern  Alaska  had  its  origin  in  Punuk  art,  what 
was  the  basis  of  the  more  northern  style,  that  prevailing  from  Norton 
Sound  to  the  Arctic  coast  and  extending  even  to  north  Greenland? 
The  archeology  of  the  Arctic  coast  of  Alaska  is  not  yet  sufficiently 
known  to  permit  a  definite  answer,  but  according  to  present  indications 
it  might  well  have  been  the  Old  Bering  Sea  style  i,  with  its  profusion 
of  straight  lines  and  spurs,  that  gave  rise  to  the  northern  style  of 
modern  art  in  which  the  spurred  line  is  such  an  important  element. 
When  we  examine  the  art  of  other  prehistoric  phases  of  Eskimo 
culture  we  find  certain  correspondences  to  individual  Old  Bering  Sea 
elements,  but  few  definite  clues  which  might  throw  light  upon  the 
origin  of  the  art  itself.  Thule  art  ofl:'ers  nothing  for  comparison,  since 
it  differs  in  no  way  from  the  northern  style  of  modern  Alaskan  art. 
The  art  of  the  Dorset  culture,  consisting  of  straight  lines,  long,  oblique 
spurs,  and  short  detached  lines,  is  somewhat  suggestive  of  Old  Bering 
Sea  style  i  (compare,  for  example,  pi.  12,  fig.  14,  pi.  13,  figs.  7,  8,  and 
pi.  19,  fig.  I,  with  Jenness,  1925,  fig,  9,  and  Mathiassen,  1927,  vol.  i, 
pi.  62,  and  vol.  2,  fig.  10,  i).  The  long,  slanting  spurs  and  detached 
lines  on  these  Dorset  objects  produce  an  effect  rather  similar  to  that 
of  Old  Bering  Sea  style  i.  The  human  faces  carved  on  some  of  the 
Dorset  specimens  show  little  resemblance  to  those  of  the  Old  Bering 
Sea  culture,  but  the  carving  of  a  mammal  head  with  ears  in  relief 
(Jenness,  1925,  fig.  7,  ;')  is  not  unlike  that  seen  on  the  knife  handle 
from  the  Hillside  site  (pi.  12,  fig.  12).  The  possible  significance  of 
these  resemblances  is  somewhat  obscured  by  the  fact  that  the  orna- 
mentation in  both  cases  is  very  simple,  and  therefore  of  questionable 
value  as  indicating  genetic  relationship. 
The  Dorset  practice  of  carving  a  crude  human  face  on  the  side  of 
an  object  finds  a  parallel  in  the  Aleutian  Islands,  where  Jochelson 
reports  finding  many  fragments  of  bone  dart  heads  decorated  in  this 
manner  (Jochelson,  1925,  p.  95,  figs.  81-83).  If  one  may  judge  from 
Jochelson's  few  illustrations,  these  Aleutian  carvings  are  quite  different 
from  the  realistic  human  faces  found  at  other  ancient  sites  in  this  part 
of  Alaska  (e.  g.,  Weyer,  1930,  figs.  13  and  23,  c — Alaska  Peninsula; 
