Am'/u°iy?i9OTarn1' ]  Alkaloidal  Assay.  303 
In  some  cases,  and  notably  when  the  alkaloidal  residue  is  dirty, 
traces  of  ammonia  adhere  to  the  extractive  even  after  evaporation 
to  almost  dryness. 
Such  then  are  the  principal  difficulties  of  the  shaking  out  process: 
the  occurrence  of  emulsions,  and  the  presence  of  ammonia. 
The  breaking  up  of  emulsions  may  be  assisted  in  numerous  ways. 
Interesting  and  ingenious  as  some  of  these  methods  are,  it  may  be 
more  advisable  to  occupy  oneself  with  the  problem  of  entirely 
preventing  their  formation. 
With  regard  to  the  elimination  of  ammonia  it  was  thought  that 
the  great  ' ammonia  capacity  of  water  compared  with  that  of  the 
volatile  solvents  might  be  taken  advantage  of,  and  the  following 
experiment  was  conducted  with  a  view  to  fixing  the  possibilities  in 
that  direction  : — 
To  a  separator  containing  15  c.c.  of  water  and  10  c.c.  of  chloro- 
form, five  drops  of  ammonia  water  were  added  and  the  mixture  was 
shaken.  'After  complete  separation  the  chloroform  was  shaken 
through  separators  2,  3,  4  and  5  in  succession,  each  of  which  con- 
tained 5  c.c.  of  water,  and  finally  rejected.  Three  subsequent  10  c.c. 
portions  of  chloroform  followed  the  same  route  in   like  manner. 
N 
For  neutralization,  separators  3,4  and  5  required  of  — H0S04,  2  5 
5° 
c.c,  04  c.c.  and  0  05  c.c.  respectively.  It  will  be  seen  that  in  the 
presence  of  the  merest  traces  of  ammonia  this  difference  in  solvent 
power  is  not  entirely  devoid  of  analytical  resources. 
The  insolubility  of  ammonium  sulphate  in  strong  alcohol  sug- 
gested precipitation  of  the  ammonia  ;  5  c.c.  of  water  containing  a 
few  centigrammes  of  this  salt  was  added  to  95  c.c.  of  absolute  alcohol. 
Part  of  the  resulting  precipitate  was  fine  enough  to  pass  through 
four  folds  of  filter  paper  specially  manufactured  to  retain  freshly 
precipitated  barium  sulphate.  Filtration  was  eventually  effected 
through  an  asbestos  packed  Gooch  crucible  by  pressure.  The  filtrate 
indicated  by  nesslerizing  o-OOOOl  gramme  NH3  per  cubic  centimeter. 
On  attempting,  however,  to  precipitate  the  ammonia  as  sulphate  from 
a  solution  of  ammonium  acetate  in  95  per  cent,  alcohol  by  means  of 
sulphuric  acid,  it  was  found  that  while  the  free  acetic  acid  scarcely 
affected  the  precipitation,  the  slightest  excess  of  sulphuric  acid  did 
so  by  rendering  the  salt  comparatively  soluble. 
It  was  thought  that  the  precipitation  of  ammonia  might  be 
accomplished  by  excess  of  tartaric  acid.    Accordingly  to  a  similar 
