NO.    I  ARCHEOLOGY    OF    ST.    LAWRENCE    ISLAND COLLINS  319 
parallel  with  the  line  holes,  and  traces  of  decoration  (Yonemura,  1935, 
fig.  5)" 
We  may  now  observe  the  distribution  of  the  toggle  harpoon  head 
in  northern  Eurasia.  First  of  all,  we  must  note  the  presence  of  old 
harpoon  heads  of  Eskimo  type  as  far  west  in  Siberia  as  the  Kolyma 
River.  Two  of  these,  along  with  fragments  of  pottery  and  stone 
blades,  were  excavated  from  old  house  ruins  by  Prof.  H.  U.  Sverdrup 
and  have  been  described  by  Mathiassen  (1927,  vol.  2,  p.  180).  Both 
have  open  sockets  with  lashing  slots,  an  irregular  spur,  and  one  promi- 
nent barb  with  an  opposite  side  blade ;  they  are  therefore  identical 
with  one  of  the  most  characteristic  Birnirk  types  (not  Thule  as  stated 
by  Mathiassen).  These  harpoon  heads,  being  specifically  Eskimo  in 
type,  afford  strong  confirmation  of  the  theory  that  the  ruins  of  under- 
ground houses  found  along  the  Arctic  coast  of  northeastern  Siberia 
are  actually  those  of  Eskimos  who  formerly  occupied  the  region.  It 
would  be  all  the  more  interesting,  therefore,  if  information  were 
available  concerning  the  ruins  of  underground  houses  reported  still 
farther  to  the  west  along  the  stretch  of  Arctic  coast  now  occupied 
by  the  Yukaghir  ( Jochelson,  1910-26,  pp.  346-348)  ;  for  if  an  Eskimo 
"  With  regard  to  these  old  harpoon  heads  from  Japan  it  should  be  mentioned 
that  a  striking  parallel  is  found  in  the  iron  toggle  heads  of  the  harpoon-arrows 
which  the  Philippine  Negritos  use  in  hunting  land  mammals.  These  have  been 
described  by  Schadenberg  (1880,  pi.  7,  figs.  8,  9)  ;  Meyer  (1893,  p.  14.  figs,  i,  2; 
pi.  6,  figs.  2,  3;  pi.  8,  fig.  2)  ;  Mason  (1902,  p.  236)  ;  and  Krieger  (1926,  pi.  4, 
figs.  II,  13).  The  similarity  between  the  two  forms  is  truly  remarkable.  Thus 
one  of  the  examples  illustrated  by  Meyer,  figure  i,  page  14,  has  the  same  tri- 
furcated,  lateral  form  of  spur  as  the  above  mentioned  Japanese  heads.  The  other 
Negrito  specimens  resemble  the  Japanese  type  illustrated  by  Kishinouye  in 
plate  22,  figure  60.  This  is  a  bladeless  head  with  a  closed  socket,  a  single, 
rather  flaring  spur,  line  hole  parallel  with  spur,  and  two  pairs  of  opposite  barbs 
between  line  hole  and  tip.  All  of  the  Philippine  examples  mentioned  have  this 
general  shape,  the  only  difference  being  that  the  line  hole  is  at  right  angles 
to  the  spur;  one  of  them  (Krieger,  pi.  4  fig.  11)  has  even  the  two  pairs  of 
barbs  between  line  hole  and  tip.  All  of  the  toggle  heads  have  closed  sockets 
which  fit  on  to  a  loose  wooden  foreshaft  very  much  like  the  Eskimo  form ; 
this  is  attached  to  the  line  by  means  of  a  cord  wrapping;  the  base  of  the  fore- 
shaft  rests  in  a  squarish  socket,  an  enlargement  of  the  shaft.  The  structural 
agreement  between  the  two  forms  is  so  close  that  one  is  led  to  suspect  that 
the  Negritos,  in  making  the  iron  heads  for  their  harpoon  arrows,  have  had  in 
mind  a  bone  prototype  like  the  Japanese  heads.  It  is  a  point  that  might  have 
some  bearing  on  the  prehistory  of  the  Negrito,  especially  in  view  of  the  strong 
probability  that  a  Negroid  element  was  present  in  the  Japanese  archipelago  in 
Neolithic  times  and  because  of  the  cultural  parallels  known  to  exist  between 
Neolithic  Japan  and  southeastern  Asia  where  the  earliest  population  seems 
to  have  been  predominantly  Negroid. 
