1030 
striictures  appear,  biit  wliere  tlie  protoplasm  lias  a marble  appearance. 
The  cells  witli  tlie  evident  neiirofibril  structure  inay  be  of  the  most 
vai-jing  sizes  and  rnaj  also  occiir  in  different  stages  of  development; 
this  is  shown,  among  other  ways,  bj  the  fact  that  in  mammals  of 
post-foetal  ages  a few  bi-polar  eells  are  fonnd  in  this  gronp,  besides 
the  ordinary  unipolar  cells,  (see  tig.  12).  Tlie  cells  that  are  without 
any  neurofibril  structure,  or  have  merely  tracés  of  this,  are  similarly 
of  very  different  sizes,  and  on  se^^eral  of  them  I have  fouud 
formations  which  conld  hardly  be  interpreted  otherwise  than  by 
assuming  that  the  cell  is  dividing  amitotically.  On  the  other  hand 
I have  not  in  a single  case  been  able  to  oliserve  any  indications  of 
amitotic  division  in  a cell  of  the  former  type,  in  which  the  nenro- 
fibril  structure  was  evident.  The  apolar  cells  also  belong  to  this 
category. 
In  a number  of  preparations  from  the  spinal  ganglia  of  young 
animals  (dogs)  1 have  fbund  colonies  of  nerve-cells  sitnated  within 
the  same  capsule.  The  number  of  cells  in  these  colonies  varied  con- 
siderably.  Fig.  4 shows  one  of  these  colonies  with  seven  cells,  in 
which  at  a few  places  protoplasmic  bridges  (bridging  fibrils)  go 
from  one  cell  to  the  other;  there  are  no  processes,  and  the  cells 
show  a pale  shade  of  colour;  there  .is  no  neuro-fibril  impregnation. 
The  references  in  liteiature  to  this  condition  and  a more  detailed 
description  of  it  will  be  given  below.  Tracés  of  this  differerice  in 
nenro-tibril  impregnation  which  is  found  in  the  spinal  ganglia  are 
also  seen  in  the  central  nerve  System,  althongh  it  is  not  so  striking  there. 
These  facts  have  led  me  to  set  up  the  following  working  hypo- 
thesis: The  afünity  of  the  neuro-tibrils  in  the  nerve-cells  to  the 
silver  salts  (reducing  power)  seems  to  vary  from  being  more  or  less 
powerful  to  total  disappearance  duriiig  certain  metabolic  or  functional 
stages.  The  majority  of  the  pale  cells  seem  thus  to  belong  to  such 
an  early  stage  of  development  that  no  neuro-fibrils  have  yet  been 
fully  developed  in  them. 
We  have  now  reached  the  hearl  of  the  problem  of  division,  viz. 
the  increase  of  the  ganglion  cells  in  the  spinal  ganglion.  With  regard 
to  this  problem  Hatai 'j  writes  as  follows;  “We  can  oidy  say  at 
present  concerning  the  division  problem  that  the  nerve  cells  in 
vertebrates,  as  well  as  in  invertebrales,  have  the  centrosome  and 
the  sphare,  which  are  regarded  as  the  dynamic  centres  of  the  mitotic 
divisions,  and,  further,  that  this  centrosome  is  able  to  take  the  first 
0 Hatai,  S.  On  the  Presence  of  the  Gentrosoine  in  Certain  Nerve  Cells  of  the 
white  rat  J.  Comp.  Neur.  Vol.  XI.  N“.  1.  1901. 
