1042 
literatiire  — I have  not  found  a single  mention  of  them  in  a whole 
series  of  recent  pnblications  on  tliis  subject  that  I have  looked 
throngh.  Muller  gives  tlie  name  oï  “Cellkolonien”  to  these  groups  of 
nerve-cells  and  distingnishes  between  regnlar  and  irregnlar  colonies. 
“Die  ersteren”  — the  regnlar  ones  — “sind  nach  anssen  durch  eine 
cirkelrnnde  Kapsel  vom  selbigen  Anssehen  wie  diejenige,  welche 
die  grossen  Zeilen  nmgiebt,  begrenzt;  innerhalb  dieser  Kapsel  fmden 
zich  zwei,  drei  oder  vier  Zeilen  sehr  regelmassig  wie  Sectoren  um 
einen  Mittelpnnkt  geordnet”.  Müi>ler  also  found  bridges  of  protoplasm 
connecting  the  dilferent  cells  of  the  colonj  with  each  other.  I have 
not  found  in  my  preparations  any  colonies  of  cells  which  showed 
this  regnlar  arrangement  of  their  cells,  resembling  a sector  of  a 
circle,  althongh  there  are  several  tigures  of  colonies  in  which  the 
cells  are  very  nearly  equal  in  size;  but  in  these  cases  they  ai‘e 
situated  side  by  side,  althongh  they  do  not  always  form  such  long 
rows  as  the  one  shown  in  fig.  4.  Most  of  the  colonies  observed  by 
me  are  quite  clearly  built  up  of  cells  that  are  different  in  size,  and  ' 
it  seems  as  if  one  might  place  them  all  in  the  gronp  that  Müller 
describes  as  irregular.  With  regard  lo  the  significance  of  these 
colonies  Müller  writes:  “Vielleicht  steht  das  Vorkommen  dieser 
Bildungen  mit  Regenerations-phanomenen  in  den  Spinalganglien  in 
Verbindung”,  but  he  points  out  that,  as  he  had  no  opportunity  of 
studying  the  processes  of  these  cells,  his  statement  on  this  point  can 
only  be  a sn|)position.  He  continues:  “So  viel  geht  jedoch  aus  dem 
unbedeutenden  Vorkommniss  bei  alteren  Thieren  von  diesen  Bildungen 
— Koloniën  und  Halbmonden  — welche  bei  jungen  Thieren  zahlreich 
auftreten,  hervor,  dass  sie  Entwicklungsstadien  von  Ganglienzellen 
reprasentieren  und  ferner,  dass  die  Entwickelung  der  Spinalganglien 
eine  langsame  ist,  welche  erst  in  spateren  Zeitraurnen  von  dem 
Leben  des  Thieres  abgesch lossen  wird.” 
In  tearing  preparations  of  older  animals  the  same  investigator  found 
that  the  crescent-shaped  cells  that  are  situated  within  the  same  ■ 
capsule  as  other  ganglion  cells,  have  no  processes.  The  observations  j 
of  mine,  however,  are  not  made  from  tearing  preparations,  in  which  ■ 
one  has  of  course  always  to  reckon  with  the  possibility  of  the 
removal  of  processes  that  have  really  been  present,  but  are  ! 
made  from  continnous  series  of  intensely  impregnated  Bielschowsky- 
preparations,  in  which  one  can  very  easily  look  for  these  colonies 
section  by  section.  In  the  series  of  sections  from  which  fig.  4 
is  taken  there  is  no  tracé  of  any  processes.  The  spinal  ganglion 
in  (piestion  is  intensely  impregnated  according  to  the  method  mentioned 
above.  The  impregnation  is  very  successful ; not  only  the  axons, 
