1028 
nated  dorsal  roots  I have  seen  some  non-medullated  fibres,  but 
they  have  never  occurred  in  my  preparations  in  as  great  niitnbers 
as  R.  seems  to  have  seen  them.  A large  nnmber  of  these  üner  fibres 
have,  as  one  finds  on  closer  investigation,  a mednllary  sheath, 
although  rather  a thin  one.  Tliis  fact  has  caused  me  to  nndertake 
a closer  and  renewed  examination  of  the  question  of  the  relation 
between  the  total  number  of  nerve  fibres  in  the  dorsal  root  and  the 
total  nnmber  of  ganglion  cells  in  the  spinal  ganglion  belonging 
to  this. 
From  the  right  side  of  a 3,5  years’  old  dog  the  spinal  ganglia 
witli  their  dorsal  roots  from  Th  VI,  L IV  and  L Vil  were  taken. 
After  fixation  and  silver-irnpregnation  (according.  to  my  modifications 
of  the  B.-method)  the  spinal  ganglia  were  put  in  unbroken  series 
of  sections,  15  ft  thick,  and  the  cross-sections  at  h and  c fig.  1 were 
made  10  ft  thick. 
The  resnlts  obtained  were  as  follows: 
Th.  VI.  Total  number  of  ganglion-cells  = 8422 
,,  ,,  ,,  nerve-fibres  at  6 = 6198 
„ „ „ „ „ c=:  6297 
L IV.  Total  number  of  ganglion  cells  =12181 
,,  ,,  ,,  nerve-fibres  at  è = 9003 
„ c=  9311 
Ij  VII.  Total  number  of  ganglion  cells  = 29621 
,,  ,,  ,,  nerve-fibres  at  6 = 23627 
„ „ „ „ „c  = 23987 
These  figures  show  that,  although  each  nerve  fibre  in  the  dorsal 
root  is  counted,  one  does  not  reach  the  total  number  of  the  spinal 
ganglion  cells.  This  is  also  the  case  if  one  counts  the  nerve  fibres 
situated  immediately  peripherally  from  the  spinal  ganglion.  The 
slightly  larger  values  of  the  latter  nerve  fibres  are  all  within  the 
limit  error  in  calculation  (±  2 per  cent).  Thus  one  cannot  conclude 
from  these  figures  that  the  nerve  fibres  which  issue  out  of  the  spinal 
ganglion  into  the  nerve  are  more  numerous  than  those  which  form 
the  dorsal  root.  It  was  noticeable  that  the  mednllary  slieaths  were 
more  powerfully  developed  at  c than  at  h fig.  1.  On  the  strenglh 
of  the  resnlts  of  my  own  investigations  I wish  to  state  that  Ranson 
goes  too  far,  and  that  the  other  writers  who  have  worked  with  the 
method  of  mednllary  sheaths  do  not  go  far  enough  in  their  con- 
clusions  with  regard  to  the  relation  between  the  number  of  ganglion 
cells  and  the  number  of  nerve  fibres  in  the  dorsal  root.  Here,  as 
in  so  many  other  cases,  the  motto  “in  medio  veritas”  applies.  Thus, 
in  spire  of  Ranson,  we  must  take  into  account  the  fact  that  the 
