122 
Estimation  of  Caseine. 
f  Am.  Jour.  Ptiarm. 
\      March,  1906. 
the  amount  of  solution  for  the  entire  1,000  samples  ranged  only  from 
2-6  to  3*2  c.c.  to  25  c.c.  milk. 
The  writers  have  since  found  that  Schlossman  (Ztschr.  physiolog. 
Ch.,  vol.  22,  p.  197)  employed  a  concentrated  alum  solution  for  the 
separation  of  caseine  from  the  other  milk  albuminoids  ;  estimating 
the  amount  of  caseines,  however,  by  submitting  the  precipitation  to 
the  Kjehldahl  process.  It  is  interesting  to  note  that  he  mentions  that 
I  to  1  5  c.c.  concentrated  alum  solution  is  necessary  for  the  complete 
precipitation  of  caseine  from  10  c.c.  milk,  thus  confirming  the 
writers'  figures  on  25  c.c.  milk,  given  above. 
Finding  the  estimation  satisfactory  in  principle,  efforts  were  made 
toward  greater  accuracy.  In  using  alum  solution  alone,  the  end  of 
precipitation  is  obscured  by  a  slight  greasiness  of  the  liquid  and 
attempts  were  made  to  obviate  this  by  addition  of  alkali,  such  as 
sodium  carbonate,  bicarbonate  or  borate.  While  these  did  aid  in 
giving  a  clearer  precipitation  their  use  complicates  final  calculations 
and  was  abandoned. 
The  next  efforts  were  toward  proving  the  end  of  reaction  by  indi- 
cations of  alum  in  the  whey,  but  work  in  this  direction  was  hampered 
by  lack  of  a  delicate  color  reagent  for  alum.  Decoction  of  logwood, 
suggested  by  Mrs.  Richards  ("  Leffmann's  Water  Analysis,"  1895, 
p.  59),  gives  a  brilliant  color  test,  but  experiments  proved  that  it 
was  not  sufficiently  delicate  for  the  dilution  of  alum  found  in  the 
whey. 
Failing  to  find  a  sensitive  color  test  of  potassa  alum,  attention 
was  turned  to  ferric  alum,  experiments  showing  that  both  ferric  and 
chrome  alum  precipitated  caseine  from  milk  with  the  same  facility 
as  does  potassa  alum. 
The  first  plan  was  to  add  a  decinormal  ferric  alum  solution  to 
diluted  milk,  filtering  a  few  drops  of  the  whey  upon  a  crucible  top  and 
bringing  in  contact  with  the  filtrate  a  drop  of  solution  of  potassium 
ferrocyanide  or  potassium  sulphocyanate.  The  result  of  these  ex- 
periments was  far  from  satisfactory,  it  being  found  that  the  end  of 
reaction  was  scarcely  more  sensitive  in  the  high  degree  of  dilution 
required  than  it  was  with  alum  and  logwood  decoction. 
An  interesting  fact  noted  was  that  the  filtrate  did  not  respond  to 
the  iron  test  with  the  reagents  until  long  after  all  the  caseine  had 
been  precipitated ;  while  on  the  other  hand,  when  the  ferrocyanide 
solution  was  first  treated  with  acetic  acid,  the  response  was  much 
