NO.     I  ARCHEOLOGY    OF    ST.    LAWRENCE    ISLAND COLLINS  3OI 
flat  circles  of  style  i.  Since  these  seem  to  have  developed  into  the 
larger,  elevated,  paired  "  eyes  "  of  styles  2  and  3,  there  might  seem 
to  be  little  need  of  bringing  Chinese  art  into  the  comparison  at  all. 
However,  regardless  of  antecedent  stages  in  either  region,  the  resem- 
blances between  the  developed  Eskimo  and  Chinese  eye  motives  are 
so  striking  that  I  cannot  avoid  the  feeling  that  we  are  dealing  with 
homologous  features.  Accordingly,  for  the  present,  we  can  only 
point  to  the  facts  which  lead  to  this  conclusion  and  leave  for  future 
determination  the  exact  nature  of  the  cultural  movements  and  con- 
tacts responsible. 
Jenness  has  called  attention  to  the  resemblances  between  Old  Bering 
Sea  and  Melanesian  art,  and,  strangely  enough,  this  is  one  of  the 
few  regions  where  we  find  anything  approaching  Old  Bering  Sea 
composition.  Melanesian  art  employs,  among  other  elements,  the 
spurred  line,  rows  of  alternating  spurs — usually  thickened  into  tri- 
angles— zigzags,  and  freehand  circles,  and  although  it  exhibits  a  far 
greater  variety  of  design  and  a  freedom  of  execution  that  contrasts 
strongly  with  the  delicacy  and  restraint  of  Old  Bering  Sea  art,  there 
are  nevertheless  certain  correspondences  to  be  observed.  The  closest 
perhaps  are  the  designs  applied  to  the  pronged  ends  of  drums  or  to 
lime  spatulas,  where,  just  as  in  Old  Bering  Sea  art,  pairs  of  circular 
or  elliptical  figures  are  so  arranged  in  relation  to  curving  lines  and 
surfaces  as  to  suggest  the  eyes  of  an  animal  (Haddon,  1912,  vol.  4, 
p.  371,  fig.  359).  Also,  the  common  Melanesian  practice  of  placing  a 
circle  or  "  eye  "  at  the  larger  end  of  a  long,  tapering,  oval  space  or 
panel  recalls  the  equally  characteristic  Old  Bering  Sea  design  of  a 
circle  placed  at  the  outer  ends  of  two  converging  lines.  Unfortunately, 
we  know  nothing  of  the  ancestry  of  Melanesian  art.  The  present  evi- 
dence suggests  a  local  origin,  as  there  seems  to  be  nothing  comparable 
to  it  on  the  Asiatic  mainland.  The  fact  that  certain  other  Oceanic  cul- 
ture elements  such  as  shouldered  celts,  star-shaped  club  heads  of  stone, 
and  other  forms  of  stone  implements  have  their  parallels  in  Neolithic 
Japan,  China,  or  Indo-China  suggests  a  stratification  of  Oceanic  cul- 
ture, such  elements  as  these  being  older  than  the  various  local  art  styles. 
The  resemblances  between  modern  Melanesian  and  Old  Bering  Sea 
art  would  therefore  be  explained  as  the  result  of  convergence,  an 
explanation  that  appears  the  more  probable  in  view  of  the  fact  that 
the  resemblances  here  are  closer  than  between  the  old  Eskimo  art  and 
any  other,  ancient  or  modern,  from  the  intervening  Asiatic  mainland. 
There  remains  to  be  considered  the  question  of  the  Old  World 
relationships  of  Punuk  art.  When  I  first  described  Punuk  art  I  as- 
sumed that  it  represented  an  outgrowth  from  that  of  the  Old  Bering 
