30  ASSAY  OF  ALKALOIDS. 
liquor,  by  filtering  and  washing, — the  latter,  together  with  all 
test-liquids  and  washings  added,  only  reaches  the  strength  of 
Groves'  original  solution  at  the  last. 
He  does  not  perceive  this,  but  impresses  upon  his  hearers  the 
belief  that  I  continue  to  dilute  the  liquid  indefinitely.  "But 
when  strong  solutions  are  used  the  precipitate  is  either  clotty 
or  gelatinous,  and  not  to  be  washed  without  much  water,  thus 
reproducing  the  difficulty  sought  to  be  avoided." 
Mr.  Groves'  logical  conclusions  sometimes  differ  from  mine, 
as  in  this  paragraph  on  morphia. 
His  theory  requires  that  equal  parts  of  both  liquids  should 
be  consumed,  but  his  reaction  stops  at  70,  not  at  100.  Hydro- 
sulphate  of  ammonia  indicates  mercury  in  solution,  and  after 
(or  before  ?)  adding  more  iodohydrargyrate,  strychnia  likewise 
gives  a  precipitate.  Hence  he  concludes  that  there  is  already 
an  excess  of  iodohydrargyrate.  So  that  out  of  70  parts  of  test- 
liquor,  which  contain  about  one  and  a  half  equivalents  of  mer- 
cury, he  has  actually  extracted  his  two  equivalents  and  over. 
He  is  bound  to  contradict  the  evidence  of  his  senses,  because 
they  seem  to  confirm  my  view  ;  nor  does  he  stop  to  inquire  and 
to  decide  by  the  simplest  test  possible  what  causes  the  precipi- 
tate to  re-dissolve.  His  "  series  of  experiments  showed  the 
cause  of  this  to  be  the  ready  solubility  of  iodohydrargyrate  of 
morphia  in  iodide  of  potassium.  It  is  also  soluble  in  hot  water." 
I  stated  more  than  a  year  ago  that  this  compound  of  morphia 
was  unusually  soluble  in  water,  meaning,  of  course,  water  of 
the  ordinary  temperature,  and  was  therefore  not  surprised  that 
on  evaporating  the  filtrate  to  one-third,  the  whole  of  the  precipi- 
tate held  in  solution  separated  in  perfect  order.  Would  it  have 
separated  if  the  iodide  of  potassium  had  been  its  solvent  ? 
There  remains,  in  fact,  no  doubt  but  that  my  original  pro- 
portion of  3  mercury  to  2  morphia  correctly  expresses  the  pro- 
portion of  test-liquor  required  for  this  alkaloid.  And  no  ex- 
periments of  mine  have  as  yet  been  able  to  induce  me  to  depart 
from  the  proportion  in  which  I  have  combined  the  ingredients, 
and  thus  to  destroy  the  uniformity  of  the  method. 
As  here,  so  elsewhere,  the  composition  and  quantity  of  the 
precipitate  are  not  always  expressed  by  the  equivalents  of  test- 
