1 62  Contributions  from  H.  M.  Gordin.  {AmAP0rnr;i9S.arm' 
We  see  that  by  Dunstan  and  Short's  method  we  only  extract 
about  83  per  cent,  of  the  total  alkaloids. 
In  order  to  select  the  best  method  of  extraction  among  the  great 
number  proposed  for  that  purpose,  the  custom  has  generally  been 
to  make  comparative  assays  of  one  and  the  same  drug  by  different 
methods  and  give  that  method  the  preference  which  gives  the 
highest  results.  This  way  of  establishing  the  correctness  of  a 
method  is  certainly  not  without  faults.  It  is  not  the  high  results  we 
want,  but  the  true  results.  Just  as  some  faulty  features  of  a  method 
are  liable  to  give  results  below  the  truth,  there  are  others  which 
might  give  results  above  the  truth.  The  explanation  of  a  method 
giving  too  low  results  is  easily  found  by  admitting  that  either  the 
extraction  is  not  complete  or  the  operations  of  the  method  involve 
a  loss  of  some  alkaloid.  The  cause  of  too  high  results  is  not  so 
easily  found,  but  a  method  should  not  be  adopted  unless  it  be  shown 
that  it  gives  exact  results,  or  at  least  approaches  the  truth  nearer 
than  any  other  method. 
Now,  it  is  certainly  possible  to  work  out  for  every  drug  a  method 
which  will  have  this  quality.  Such  a  method  might  be  too  tedious,, 
too  expensive,  and  too  complicated  for  general  use.  But  such  a 
method  could  be  used  as  a  standard  with  which  simpler  and  quicker 
methods  could  be  compared.  If  it  can  be  shown  that  the  standard 
method  gives  the  most  exact  results  obtainable  at  the  present  state 
of  our  knowledge  and  that  among  the  many  simpler  methods  a 
particular  one  gives  results  which  approach  those  obtained  by  the 
standard  method  better  than  any  other  method,  that  particular 
method  should  be  chosen  for  general  use.  Any  other  method  giv- 
ing either  higher  or  lower  results  than  the  standard  method  should 
be  rejected.  How  to  find  such  a  standard  method  I  shall  try  to 
show  in  the  case  of  a  few  drugs  only,  but  the  principles  can  be 
extended  to  any  other  drug. 
The  necessary  and  sufficient  demands  which  should  be  put  to  a 
standard  method  are  :  ^ 
(1)  That  the  exhaustion  should  be  so  complete  that  no  alkaloid 
could  be  found  in  about  5  grammes  of  the  dregs  by  the  method 
explained  above. 
(2)  The  operations  involved  in  the  standard  method  should  only 
be  such  as  are  not  liable  to  injure  the  alkaloid  under  consideration. 
Heat,  strong  acids  or  strong  alkali  and  prolonged  exposure  to  the 
