1 089 
tliis  figure  as  a spiiial  ganglion  oell  at  sncli  an  earlj  stage  of  deve- 
lopment  that  it  had  not  lost  lts  powei'  of  increasing  in  nnmber 
throLigh  mitotical  division.  I fonnd  anoibei'  mitosis  of  tliis  kind  in 
the  series  jnst  mentioned.  Among  the  other  gronp  of  mitoses, 
namely  those  in  cel  Is  that  are  situated  inside  or  in  the  capsule  of 
an  older  ganglion  cell,  mj  preparation  shows  at  least  a few  forrns 
in  which  one  can  clearlj  follow  the  capsule  peripherall}^  of  the  cell 
that  is  engaged  in  mitotic  division  and  where  the  latter  must  therefore 
be  situated  beneath  the  capsule.  There  are  thus  good  reasons  to 
support  the  assumption  that,  even  among  this  gronp  of  mitoses, 
sorne  are  to  be  referred  to  very  joung  undifferentiated  cells,  which 
on  good  grounds  — tbr  instance  on  account  of  their  position  — 
can  be  assumed  to  develop  into  Jierve  cells.  Bj  far  the  larger 
number  of  mitoses  are,  however,  undoubtedly  to  be  referred  to 
o.rdinarj'  capsular  cells.  But  is  the  ditference  between  the  capsular 
cells  and  the  nerve-cells  really  so  great?  Are  not  the  former  perhaps 
to  be  regarded  as  matrix  cells  for  the  latter?  I must  leave  these 
problems  to  a subsequent  and  more  detailed  account  of  this  question 
and  confme  myself  to  saying  that  there  are  points  in  the  prepa- 
ration that  support  siich  an  assumption ').  These  facts  are  all  the  more 
worthy  of  attention  because,  among  the  in vestigators  who  formerlj' 
looked  for  mitoses  in  spinal  ganglia,  Fi.emming  ’),  Daal  and  Len- 
HOssEK  have  been  unable  to  show  any  in  young  animals.  Müller  ®), 
on  the  other  hand,  found  them  in  new-born  animals,  but  in  no 
later  age.  The  very  large  number  of  mitoses  in  the  spinal  ganglia 
shown  in  the  present  and  other  investigations  of  young  animals 
clearly  support  the  considerable  post-embryouic  increase  in  the 
number  of  capsular  cells  in  this  region,  an  increase  that  could 
scarcely  be  explained  if  the  ganglia  did  not  increase  in  number 
too.  In  my  opinion  the  exceedingly  great  number  of  mitoses  that 
are  found  in  the  spinal  nerve-cells,  according  to  what  has  been 
shown  above,  cannot  possibly  be  explained  by  an  increase  in  size 
merely  of  those  spinal  nerve-cells  which  were  already  present  at 
birth.  This  is  the  less  probable  because  the  spinal  ganglion-cells 
must  decrease  in  number  with  the  years  if  new  ones  do  not  grow 
out  and  replace  all  those  that  degenerate  and  die  away  duting 
post-embryonic  life.  And  this  degeneration  of  the  nerve-cells  is 
admitted  and  shown  by  all  the  chief  in  vestigators  of  this  problem. 
1)  See  addendum! 
Flemming,  Daal  and  Eeehossek.  Quoted  from  Müller  E. 
Müller  Erik,  Untersuchungen  über  den  Bau  der  Spinalganglien.  Nord.  ined. 
Ark.  Stockholm.  Bd.  23.  1891. 
