Am.  Jour.  Ph* 
March,  1906. 
Estimation  of  Caseine. 
127 
4b 
i6'o 
4-0 
4'i 
4'2 
4"o 
4'2 
4'2 
4'4 
4*3 
4'i 
25-0 
25 '5 
26*0 
25'3 
25-0 
25*9 
25  "5 
24'5 
25-0 
24-3 
25"5 
0-0350 
0*0357 
0-0364 
0-0354 
0*0350 
0*0362 
0-0357 
0-0343 
0-0350 
0-0340 
0-0357 
2*19 
2-23 
2*27 
2-23 
2*27 
2'33 
2*25 
2-23 
2-31 
2-28 
2-ig 
2-23 
2-17 
2'27 
6  a 
6b 
6  c 
6d 
4  c 
4d 
5  a 
5  b 
5  c 
5d 
i5'9 
15-8 
i6'o 
15-3 
15-8 
i5'6 
I5-7 
i5'9 
l6'2 
3-  8 
4-  2 
2*19 
2*26 
2-23 
2-14 
2T9 
2"I2 
15-8 
2-23 
To  establish  the  definite  standard,  it  will  of  course  be  necessary  to 
make  a  large  number  of  comparative  estimations,  and  that  with 
officially  standardized  measures  instead  of  ordinary  laboratory  glass, 
ware.  Such  work  will  be  performed  by  the  writers  during  the  com- 
ing winter,  and  it  is  hoped  that  other  investigators  will  take  up  the 
same  line  of  experiments  ;  for  if  the  ferric  alum  estimation  is  entirely 
feasible,  it  will  prove  vastly  superior  to  the  cumbersome  and  tedious 
Kjehldahl  process  now  in  vogue. 
In  conclusion,  a  few  remarks  as  to  methods  of  manipulation  may 
be  in  place.  The  most  convenient  amount  of  reacting  ingredients 
is  10  c.c.  milk  and  20  c.c.  decinormal  ferric  alum  solution  (48-1 
gramme  to  the  liter).  This  amount  of  ferric  alum  solution  requires 
exactly  20  c.c.  decinormal  thiosulphate  solution  to  decolorize  the 
iodine  liberated  by  it  and  the  acid  from  the  potassium  iodide,  and 
each  cubic  centimeter  of  thiosulphate  solution  less  than  twenty, 
required  after  combination  with  milk,  means  a  cubic  centimeter  of 
ferric  alum  solution  employed  in  precipitation. 
The  milk  and  ferric  alum  solution  are  mixed  cold  and  it  seems  to 
make  but  little  difference  whether  the  milk  be  added  to  the  ferric 
alum,  solution  or  vice  versa.  The  mixture  is  allowed  to  stand  fifteen 
or  thirty  minutes  and  then  filtered,  and  for  this  purpose  a  pledget  of 
cotton  in  a  glass  funnel  is  found  most  satisfactory.  If  filter  paper 
is  used  the  iron  creeps  to  the  edge  of  the  paper,  from  which  it  is 
washed  with  much  difficulty.  Pressure  filtration  was  tried — to 
quicken  process — but  with  results  scarcely  as  satisfactory  as  filtra- 
tion through  cotton.  At  its  best,  the  process  of  filtration  is  tedious, 
yet  we  have  washed  out  precipitates  in  two  hours  and  a  batch  started 
in  the  afternoon  can  always  be  ready  for  titration  the  next  morning. 
In  filtering,  best  results  are  obtained  by  throwing  precipitate  on 
filter  in  fairly  concentrated  liquid  :  that  is,  the  total  fluid  need  not 
