NO.    I  ARCHEOLOGY    OF    ST.    LAWRENCE    ISLAND COLLINS  1 37 
5,  The  decoration,  consisting  of  two  short  and  two  longer  Hnes, 
separated  by  two  groups  of  short  transverse  Hnes,  is  very  similar 
to  that  in  plate  34,  figure  2 ;  it  is  present  on  one  side  only.  Plate  34, 
figure  12  (cut  22,  30  in.),  of  bone,  has  the  lower  end  cut  away  to 
a  shoulder  just  above  the  tang.  There  were  three  barbs,  near  the  tip, 
so  small  as  to  be  hardly  more  than  notches.  Plate  34,  figure  13,  from 
cut  9,  depth  24  inches,  is  an  ivory  point  set  in  the  end  of  a  wooden 
shaft,  with  the  baleen  wrapping  still  preserved.  In  shape  the  point 
is  more  like  a  foreshaft  than  anything  else,  although  it  has  no  line 
hole.  On  the  other  hand  the  manner  of  hafting  is  that  of  an  arrow, 
which  it  probably  was,  in  spite  of  the  blunt  end ;  it  may  have  been  a 
child's  arrow,  never  intended  for  serious  use.  With  this  was  found 
a  "  cache  "  of  finished  and  unfinished  projectile  points,  foreshafts, 
etc.,  including  several  arrow  points  which  were  triangular  in  cross- 
section.  This  is  a  recent  type  on  St.  Lawrence  Island,  and  its  occur- 
rence, together  with  a  quantity  of  other  ivory  artifacts,  mostly  un- 
finished, at  this  one  place  in  the  midden,  but  not  elsewhere,  suggests 
that  the  entire  "  cache  "  may  have  been  later  than  the  other  material 
in  this  part  of  the  midden. 
None  of  the  tangs  have  knobs  or  projections ;  they  are  either  per- 
fectly smooth  or  merely  roughened  by  hacking.  It  should  also  be  noted 
that  all  of  the  blade  slits,  or  sockets,  are  around  4  mm  wide,  evidently 
in  order  to  accommodate  a  rather  thick,  chipped  stone  blade  like  that 
of  plate  34,  figure  i. 
We  may  now  consider  the  relative  ages  of  the  types.  It  will  be  noted 
that  five  of  the  type  5  and  type  6  specimens  come  from  the  later,  north- 
western section  of  the  midden,  and  eight  from  the  upper  levels  of  the 
older,  southeastern  section.  On  the  other  hand  all  10  of  the  examples 
of  types  2,  2  a,  3,  and  4,  in  addition  to  plate  34,  figure  11,  came  from 
the  older  part  of  the  midden,  most  of  them  from  around  the  3-foot 
level  with  the  two  examples  of  type  2  a  coming  from  depths  of  5 
and  6  feet.  It  seems  clear,  therefore,  that  types  2,  2  a,  3,  and  4  are 
older  than  types  5  and  6.  As  pointed  out  above,  the  principal  distinc- 
tion between  the  two  groups  is  the  unusual  appearance  of  the  barbs : 
in  the  older  types  they  lie  usually  close  to  the  shank,  with  a  deep  line, 
extending  upward,  and  opposite,  a  similar  line  which  may  or  may  not 
terminate  in  a  barb.  It  will  be  recalled  that  both  of  the  barbed  arrow 
points  from  the  Hillside  site  (pi.  29,  fig.  6)  had  barbs  made  in  this  way. 
Arrow  shafts. — The  lower  ends  of  two  arrow  shafts  are  shown  in 
plate  57,  figs.  4,  5  (cut  19,  32  in.,  and  cut  18,  66  in.).  Neither  of 
them  shows  evidence  of  having  been  feathered,  although  another 
specimen,  from  the  Hillside  site,  has  two  splits  extending  through 
the  shaft,  evidently  for  this  purpose. 
