364 
Estimation  of  Proteids. 
Am  Jour.  Pharm. 
July,  1889. 
precipitates  all  proteids  but  peptone,  which,  if  present,  may  be  identi- 
fied in  the  filtrate  by  the  biuret  reaction  (Kiihne),  or  better  still  by  the 
precipitate  it  gives  with  tannic  acid.  Peptone  is,  however,  absent  in 
milk,  and  also  in  colostrum,  and  in  whey ;  the  whey-proteid  of  Ham- 
marsten  is  thus  not  of  the  nature  of  peptone.  It  is,  however,  present  in 
kephir,  and  in  "  long-milk  "  (a  preparation  made  from  milk  in  Upper 
Scandinavia).  Other  portions  of  milk  were  precipitated  with  tan- 
nic acid,  and  the  filtrate  was  found  to  contain  only  such  small 
quantities  of  nitrogen  as  would  be  accounted  for  by  the  urea,  hypo- 
xanthine,  and  other  non-proteid  constituents ;  albumoses  are  absent 
therefore. 
The  method  of  estimating  the  total  nitrogen  in  the  precipitate  pro- 
duced by  tannic  acid,  and  multiplying  this  by  6 '3 7  to  obtain  the  total 
proteids,  is  recommended  as  the  most  accurate  method  in  quantitative 
investigations  of  proteids,  and  is  stated  to  produce  less  error  than  the 
more  usual  methods  involving  the  washing,  drying,  weighing  and  in- 
cineration of  proteid  precipitates. 
In  order  to  estimate  the  casein  and  lactalbuinin  separately,  rennet 
and  various  salts  added  to  saturation  have  been  recommended  ;  these 
were  all  tested,  and  magnesium  sulphate  was  found  to  be  the  best :  it 
precipitates  all  the  casein  and  the  trace  of  lactoglobulin  which  is  seem- 
ingly disregarded,  and  leaves  all  the  albumin  in  solution.  Sodium 
chloride  was  also  found  to  precipitate  all  the  casein.  The  albumin 
has  been  estimated  by  others,  by  weighing  the  precipitate  produced  by 
boiling  after  the  separation  of  the  casein.  But  boiling  was  found  to 
be  a  very  incomplete  method  of  precipitating  it ;  much  nitrogen  oc- 
curring in  the  filtrate  after  its  removal.  It  is  suggested  that  boiling 
lactalbuinin  and  other  proteids  split  them  (as  Hammarsten  found  with 
fibrinogen)  into  an  insoluble  proteid  which  is  precipitated,  and  a 
soluble  one  which  is  left  in  the  filtrate.  It  was  found  that  solutions 
of  pure  casein  when  heated  do  not  coagulate,  but  become  opalescent, 
becoming  clear  again,  however,  on  cooling.  With  regard  to  colos- 
trum, many  contradictory  statements  are  quoted  as  to  the  relative 
amount  of  casein  and  albumin.  By  the  present  method  it  is  found 
that  the  amount  of  these  two  proteids  is  very  variable  ;  that  the  quan- 
tity of  globulin  is  very  considerable,  and  that  the  non-proteid  nitro- 
genous constituents  are  more  abundant  than  in  milk. 
The  following  method  was  adopted  in  the  analysis  :  (1.)  Total  nitro- 
gen first  estimated  in  the  colostrum.    (2.)  Proteid  nitrogen  estimated 
