228  The  Structure  of  the  Starch  Grain.  {Am-M*y;im&TW' 
die  meiste^  feste  Substanz  enthalten."  Here  again  it  is  unfortunate 
that  the  descriptions  do  not  enable  us  to  determine  which  layers 
are  meant.  It  is  likely,  however,  that  the  layers  described  as 
u  schattirten  Schichten "  are  the  colloidal  layers,  and  those  desig- 
nated as  "  hellen  Schichten  "  are  the  crystalloidal  layers. 
In  summing  up  the  observations  herewith  presented,  we  find  that 
the  starch  grain  consists  of  colloidal  and  crystalloidal  substances, 
these  being  arranged  for  the  most  part  in  distinct  and  separate 
lamellae,  that  is,  at  the  point  of  origin  of  growth,  and  in  the  alter- 
nate lamellae  the  colloidal  substance  preponderates,  associated  with 
the  crystalloid  cellulose ;  whereas  in  the  other  layers  the  crystal- 
loidal substance,  consisting  for  the  most  part  of  granulose,  occurs 
in  greater  proportion. 
As  a  further  evidence  of  the  presence  of  these  crystalloidal  and 
colloidal  areas  we  may  say  that  the  peculiar  behavior  of  the  col- 
loidal layers  toward  aniline  stains  is  analogous  to  the  behavior  of 
a  section  containing  mucilage  cells  towards  these  dyes,  the  latter 
being  taken  up  by  the  mucilage  cells  alone.  Furthermore,  as  the 
characteristics  of  mucilage  cells  are  most  pronounced  in  anhydrous 
media,  as  concentrated  glycerin,  so  a  similar  effect  is  observed  in 
the  starch  grain,  but  owing  to  the  action  of  the  glycerin  in  readily 
forming  soluble  starch,  the  preparations  of  tne  starch  grain,  like 
that  of  the  cell  wall,  when  thus  stained  are  best  preserved  in  Canada 
balsam. 
The  crystalloidal  character  of  certain  of  the  lamellae  as  observed 
in  connection  with  the  swelling  of  the  grain  might  be  considered  to 
be  in  the  nature  of  microscopic  clefts,  but  it  should  be  said  that  we 
know  of  no  colloid  that  behaves  in  this  manner,  and  such  an 
assumption  does  not  seem  to  be  well  founded.  Furthermore,  this 
appearance  (crystalloidal)  might  be  considered  as  due  to  an  internal 
folding  of  the  substance  of  the  lamellae,  but  this  would  only  arise 
in  case  of  a  contraction  or  reduction  in  the  area  of  the  lamellae  ;  but 
as  we  have  already  seen,  the  grain  is  swollen  from  two  to  four  times 
its  original  size,  and  even  in  the  fragments  of  the  disintegrating 
grains  the  crystalloidal  character  is  pronounced.  It  seems  more 
likely,  as  I  have  already  pointed  out,  that  the  reason  this  structure  is 
not  apparent  under  natural  conditions  is  because  the  refractive  prop- 
erties of  the  crystalloidal  substance  so  nearly  resemble  those  of  the 
associated  colloid.    The  use  of  certain  reagents,  however,  which  are 
