NO.    I  ARCHEOLOGY    OF    ST.    LAWRENCE    ISLAND — COLLINS  307 
nient  from  the  Old  Bering  Sea  period  down  to  the  present  time.  We 
shall  now  consider  the  harpoon  heads  from  other  parts  of  the  Eskimo 
area  in  an  attempt  to  determine  the  relation  they  may  bear  to  the  St. 
Lawrence  forms,  and  if  possible  the  lines  of  development  they  may 
liave  followed. 
The  harpoon  heads  obtained  by  Jenness  at  Bering  Strait  are  in 
general  conformable  to  those  from  Gambell,  and  the  chronological 
relationships,  as  originally  outlined  by  Jenness,  find  ample  confirmation 
in  the  Gambell  sequence.  There  seems  to  have  been  a  parallel  course 
of  development  at  the  two  places,  although  the  types  were  never 
identical ;  thus  while  the  greater  number  of  the  Wales  and  Diomede 
heads  would  be  included  by  definition  in  certain  of  our  Gambell  types, 
there  are,  as  might  be  expected,  subtle  differences  in  shape,  propor- 
tion of  parts,  and  in  various  other  minor  features  which  stamp  them 
with  an  individuality  of  their  own.  Jenness  found  the  oldest  types 
to  be  those  with  trifurcated  symmetrical  spurs,  double  line  holes,  and 
side  blades — features  which  were  also  characteristic  of  the  oldest 
Gambell  types,  open  socket  types  I  x  and  I  y.  The  only  other  locality 
from  which  this  type  of  harpoon  head  has  been  reported  is  Point 
Barrow  (Wissler,  1916,  fig.  9).  Jenness  also  illustrates  a  harpoon 
head  corresponding  to  our  open  socket  type  II  y,  which  differs  from 
I  y  in  having  a  single  line  hole  and  a  spur  that  is  asymmetrical  and 
somewhat  lateral  instead  of  symmetrical  and  median  (Jenness,  1928  a, 
pi.  12,  d).  This  type,  which  appears  to  have  been  derived  from  type 
I  y,  is  of  particular  interest  because  of  its  close  relationship  to  the 
Birnirk  group,  characteristic  of  the  older  sites  around  Barrow.  The 
type  II  y  heads  from  St.  Lawrence  Island  all  have  two  side  blades ; 
some  of  the  Birnirk  heads  also  have  two,  but  the  most  common  form  is 
that  with  one  side  blade  and  an  opposite  barb.  The  latter  has  been 
found  as  far  west  as  the  Kolyma  River  in  Siberia  (Mathiassen,  1927, 
vol.  2,  p.  180)  and  seems  to  be  the  form  which  gave  rise  to  the  Thule 
type  2,  with  two  barbs. 
Although  some  of  the  earliest  St.  Lawrence  forms  have  divided 
spurs  and  side  blades — features  especially  characteristic  of  the  Birnirk 
heads — it  should  be  noted  that  heads  closely  resembling  the  Birnirk 
type  in  shape  as  well  as  in  structural  features  do  not  appear  on  St. 
Lawrence  until  the  Punuk  stage.  Some  of  these  Birnirk-like  heads 
have  been  shown  in  plate  24,  figures  9,  10;  plate  27,  figures  1-4;  plate 
28,  figures  I,  2;  plate  70,  figures  1-3.  The  fact  that  almost  all  of 
them  are  made  of  bone,  which  is  very  rarely  used  for  other  types  of 
harpoon  heads  on  St.  Lawrence,  but  which  is  the  material  from  which 
the   Birnirk  heads   are   commonlv   made,   is    further   evidence   of   a 
