1022 
provided  with  an  oil  imiiiersiuii  (Vu  a)  and  ücular  fonr.  Tlie 
preparations  were  investigated  in  the  most  carefnl  waj  step  step, 
but  noi  a single  example  ot‘  a division  of  tlie  nerve  fibres  conld  be 
discovered,  either  in  the  roots  or  in  the  continuation  of  the  nerve 
fibres  in  the  vent  ral  horn  throngh  the  spinal  oord  np  to  their  root 
cells.  I observed  a few  oases  of  spinal  ganglion-oells  which  had  the 
claviform  prooesses  mentioned  above.  These  olaviforra  formations 
were,  however,  always  within  the  oapsides.  In  no  case,  however, 
was  I able  to  discover  anything  that  conld  be  interpreted  as  a 
division  within  the  sjiinal  ganglion  of  eithei'  the  central  axon  or 
that  which  passes  peri|)herally.  I iried  to  test  the  negative  results 
obtained  from  this  investigation  in  another  way,  in  order  to  obtain 
if  possible  a positive  result.  I made  cross-sections  of  the  silver 
impregnated  malerial  throngh  the  corresponding  nerve  roots  on  tlie 
Fig.  1.  Schematic  representation  of  the  neural  growth  in  the 
dorsal  and  ventral  roots  of  the  spinal  nerves.  (a)  Cross  section  at 
the  place  with  the  smaller  and  [b)  cross  section  at  the  place  with  the 
larger  numher  of  nerve  fibres  in  the  dorsal  root.  («i)  Cross  section 
at  the  place  with  the  larger  and  {bi)  cross  section  at  the  place  with 
the  smaller  numher  of  nerve  fibres  in  the  ventral  root  at  a,  spinal 
nerve  from  a young  animal.  (ci  Cross  section  peripherally  of  the 
spinal  ganglion,  (d)  Spinal  ganglion  cell.  (?)  Ventral  root  cell. 
other  side,  some  close  to  the  spinal  cord  {a  and  a^;  lig.  J)  and 
others  close  lo,  bnt  centrally  of,  the  spinal  ganglion  (/>  and  ; fig.  1). 
As  (he  nerve  roots  that  I iiivestigated  belonged  to  the  lumbal  and 
sacral  nerves,  which  take  part  in  forming  the  qiiada  equina,  the 
distance  between  the  two  cross  sections  was  fairly  great.  The  nerve 
fibres  in  the  sectioned  |)re|)aratioii  were  counted.  There  are  three 
possibilities  for  the  totals  that  we  might  expect  to  obtain  for  the 
iinmbers  of  nerve  libres.  If  we  take  a;  «i ; b and  fig.  1 to  denote 
the  mimber  of  nerve  libres,  then  a=b  or  else  a'^b  or  finally 
a c^b.  If  a = 6 then  one  conld  scarcely  expect  any  appearance  of 
axon-division  or  any  figures  of  growth  in  the  piece  ab\  if,  on  the 
