334  SMITHSONIAN    MISCELLANEOUS    COLLECTIONS  VOL.    96 
Most  of  the  Old  Bering  Sea  and  Punuk  heads  have  had  the  stone 
blade  lashed  on  instead  of  set  in  a  socket  (pi.  46,  figs.  3-5;  pi.  60, 
fig.  2;  pi.  78,  figs.  17,  18,  20,  21).  They  are  carefully  made  in  every 
feature :  a  smooth  concavity  for  the  blade ;  an  opposite  lip  to  hold  the 
lashing  on ;  an  oval  or  rectangular  socket  for  the  end  of  the  handle ; 
and  a  groove,  knobs,  or  "  ears  "  for  the  handle  lashing.  The  "  shoe- 
shaped  "  adz  heads  of  the  Punuk  period  (pi.  78,  figs.  20,  21)  are  ap- 
parently a  local  form,  a  specialized  development  from  the  flatter  types 
such  as  plate  78.  figures  17,  18,  and  plate  46,  figure  4. 
A  less  common  Old  Bering  Sea  type  was  that  with  a  socket  at  the 
end  for  an  inserted  blade  (pi.  46,  figs.  6,  7).  In  this  respect  these 
resemble  the  modern  Alaskan  type,  which  has  the  blade  set  in  the  end, 
but  to  obtain  a  really  close  parallel  to  the  modern  Alaskan  type  we 
must  turn  to  one  of  the  latest  Gambell  specimens  (pi.  78,  fig.  19). 
The  ad^  heads  of  the  Thule  culture,  however,  are  even  closer  to  the 
modern  Alaskan  forms,  most  of  them  being  exactly  similar  in  shape, 
with  the  same  inserted  blade  and  the  same  methods  of  handle  attach- 
ment as  the  examples  illustrated  by  Murdoch  (cf.  Mathiassen,  1927, 
vol.  I,  pi.  20,  with  ^Murdoch,  figs.  133-138).  Until  we  know  what 
kind  of  adz  head  was  characteristic  of  the  Birnirk  culture,  the  his- 
tory of  this  type  must  remain  in  doubt.  Perhaps,  as  de  Laguna  has 
suggested  (  1934,  p.  173) ,  the  modern  Barrow  type  is  a  late  importation 
brought  into  Alaska  by  a  back  wash  of  Thule  culture. 
Stone  Implements 
One  of  the  most  striking  differences  between  the  two  culture  stages 
on  St.  Lawrence  Island  is  the  extensive  use  of  chipped  stone  imple- 
ments during  the  Old  Bering  Sea  period  and  their  scarcity  in  the 
Punuk.  At  the  Hillside  site  242  artifacts  of  chipped  slate,  chert, 
jasper,  chalcedony,  etc.,  were  found,  in  addition  to  large  quantities 
of  rejectage.  The  rubbing  technique  was  also  fully  developed,  for 
140  artifacts  and  innumerable  fragments  of  rul)bed  slate  were  found 
at  the  Hillside  site.  In  contrast  to  this  condition,  implements  of  chipped 
stone  were  rarely  found  at  the  three  later  sites  and  were  only  fairly 
common  at  Miyowagh.  At  these  sites  rubbed  slate  was  used  almost 
entirely.  Since  all  of  the  other  sites  are  within  three-quarters  of  a 
mile  of  the  Hillside  site,  it  is  obvious  that  the  rarity  of  chipped  imple- 
ments at  the  later  sites  is  a  strictly  cultural  phenomenon  and  not  the 
result  of  a  lack  of  suitable  material. 
When  we  turn  to  South  Alaska  we  observe  a  similar  condition, 
archeological  investigations  having  shown  that  the  older  culture  layers 
