3IO  SMITHSONIAN    MISCELLANEOUS    COLLECTIONS  VOL.    96 
present  on  any  of  the  older  harpoon  heads  from  St.  Lawrence  Island, 
Bering  Strait,  or  Barrow,  although  the  modern  iron-bladed  heads 
almost  always  have  them.  In  Alaska,  apparently,  rivet  holes  are  always 
associated  with  metal  blades,  whereas  in  the  Thule  culture  they  are 
found  on  both  iron-  and  stone-bladed  heads.  It  seems  possible  that 
the  Thule  custom  of  having  a  rivet  to  hold  stone  blades  in  place  in 
harpoon  heads,  which  is  not  actually  necessary,  may  have  had  it  origin 
in  the  Thule  lance  heads  (Mathiassen,  1927,  vol.  i,  pi.  12)  where  the 
stone  blade  was  lashed  to  the  end,  on  the  outside,  and  held  more 
securely  in  place  by  means  of  one  or  more  pegs  or  rivets. 
We  may  now  examine  the  Thule  types  as  a  whole  and  see  what 
relation  they  may  bear  to  the  Old  Bering  Sea  and  Punuk  types. 
Mathiassen  (1927,  vol.  2,  p.  11)  says  of  his  harpoon  head  classifica- 
tion :  "  This  system  is  so  arranged  that  the  forms  of  harpoon  heads 
which,  from  a  typological  point  of  view,  must  be  regarded  as  being 
the  simplest,  are  always  placed  before  the  more  complicated,  derived 
forms."  Then,  after  observing  the  distribution  of  the  several  types 
at  the  Central  Thule  sites,  he  finds  (1927,  vol.  2,  p.  14)  that  "  The 
composition  of  these  finds  indicates  that  the  forms  which  for  typologi- 
cal reasons  are  regarded  as  being  the  most  simple  are  really  the  oldest 
too.  The  Thule  group  A  i  [the  open  socket  group]  is  the  oldest  and, 
within  this,  the  fundamental  types  a,  b  i  and  c  i  are  the  oldest.  Within 
B  [the  flat  group]  it  is  I  [those  with  open  sockets]  that  is  the  oldest 
group ;  B  II  [flat  with  closed  socket]  must  be  regarded  as  the  youngest 
of  all  four  main  groups." 
As  stated  above,  the  small  bladeless  harpoon  head — Thule  type  i^ — - 
is  one  of  the  very  latest  of  the  prehistoric  types  at  Gambell,  however 
old  it  may  be  elsewhere.  It  appears  suddenly  at  Seklowaghyaget  and 
may  therefore  have  been  imported.  If  it  developed  locally  the  most 
likely  prototype  would  be  one  of  the  Birnirk-like  forms  such  as  open 
socket  type  II  (a)  y  or  II  (d)  (pi.  28,  figs.  2,  3),  but  of  this  there  is 
no  definite  evidence.  The  Thule  type  i  is  a  very  simple  form  of 
harpoon  head,  and  one  which  on  theoretical  grounds  might  well  be 
very  ancient.  But  simplicity  of  form  alone  is  of  little  value  as  evidence 
of  antiquity  until  the  possibility  of  the  alternative  explanation — de- 
generation— has  been  removed.  In  the  present  case,  lacking  the  spe- 
cific data  that  might  clarify  the  problem,  we  must  judge  the  age  of  the 
Thule  type  i  harpoon  head  from  its  setting:  if  the  accompanying 
elements  of  the  Thule  culture  prove  to  be  ancient  we  may  assume  that 
the  simplicity  of  this  particular  element  represents  an  original  condi- 
tion; if  not,  then  we  must  recognize  that  the  simplicity  in  question 
may  just  as  readily  be  the  result  of  degeneration. 
