ASSAY  OF  ALKALOIDS. 
11 
liquid  for  a  given  quantity  of  alkaloid,  but  by  a  definite  fraction 
of  that  equivalent. 
As  proof  of  this  I  will  give  some  analyses  by  other  chemists, 
while  the  main  question  shall  be  proved  by  Mr.  Groves'  testi- 
mony. 
Boedecker  (Ann.  Chem.  Pharm.  lxxiii.  p.  372 ;  Gmelin,  vol. 
xiv.  228)  describes  a  compound  of  iodide  of  mercury  with  hy- 
+ 
driodate  of  nicotia  of  the  formula,  Nic,  2HI+2HgI.  This  is 
the  composition  of  the  precipitate  formed  by  four  equivalents  of 
mercury,  but  crystallized  from  excess  of  hydriodic  acid. 
Wertheim  (Journal  f.  pr.  Chem.  vol.  91,  p.  481 — Chem.  Cen- 
tralblatt,  1864,  p.  791),  gives  the  preparation  of  another  iodo- 
hydrargyrate  of  nicotia  of  the  formula — Nic+HI-f2HgI,  ob- 
+ 
tained  by  boiling  an  alcoholic  solution  of  Nic  HI,  with  a  large 
excess  of  mercuric  oxyd.  The  same  chemist,  in  the  same  paper, 
describes  a  third  compound,  without  hydriodic  acid,  produced 
by  treating  a  mixture  of  mercuric  iodide  in  nicotia  with  boiling 
+ 
water  ;  this  has  a  composition  according  to  Nic+2  Hgl. 
Which  of  these  formulae  has  Mr.  Groves  found  by  analysis  ? 
He  unhesitatingly  adopts  the  second,  while  in  his  volumetric  as- 
say he  had  to  use  almost  exactly  the  quantity  of  mercury  which 
my  assay  requires,  i.  e.,  four  equivalents.  What  then  has  be- 
come of  the  two  equivalents  of  mercury  ? 
Now,  one  of  the  very  first  points  upon  which  Mr.  Groves 
differs  with  me  is  this  excess  of  mercury  in  solution,  of  which, 
in  my  paper  of  December  10th,  1862,  I  spoke  in  the  main  as 
follows  :  "  In  the  reactions  between  some  of  the  alkaloids  and 
the  iodohydrargyrate  of  potassium  (the  solution  of  one  HgCl 
in  3KI)  the  mercury  partly  remains  in  solution.  For  this  rea- 
son not  a  solution  of  iodide  of  mercury  must  be  used,  inasmuch 
as  with  a  solution  of  the  latter  the  results  very  much  differ  ; 
nor  must  the  solution  of  alkaloid  be  added  to  the  mercuric  so- 
lution, but  the  latter  to  the  former."  This,  I  presume,  shows 
that  even  then  I  was  aware  of  several  causes  of  ambiguity,  and 
that  in  order  to  secure  uniformity  in  the  results,  I  desired  to 
use  but  one  test  liquid  and  one  mode  of  proceeding. 
Mr.  Groves,  on  the  other  hand,  concludes  that  my  mode  of 
