1 66  Contributions  from  H.  M.  Gordin.  {A%Jp7if;iSJi*rm- 
two  layers,  25  c.c.  of  the  upper  layer  was  removed  into  a  tared  alu- 
minum dish.  After  complete  evaporation,  drying  at  1250  C.  (to 
remove  all  trace  of  chloroform),  and  weighing,  0  0564  gramme  of 
strychnine  was  found  to  have  been  contained  in  the  25  c.c.  ether- 
chloroform.  This  is  more  than  half  the  amount  of  strychnine 
taken  by  0-0041  gramme,  or  nearly  8  per  cent,  in  excess. 
As  can  be  seen  from  these  experiments,  the  error  in  excess  is  far 
greater  than  the  error  causing  a  lowering  of  the  results.  It  is  for 
this  reason  that  Keller's  method  (with  the  addition  of  water)  gen- 
erally gives  higher  results  than  when  the  water  is  left  out.  This 
was  proved  by  the  following  experiment : 
(1)  Fifteen  grammes  belladonna  root  in  No.  80  powder  were 
digested  with  150  grammes  of  a  mixture  of  three  parts  ether  and 
one  part  chloroform  for  ten  minutes.  Ten  c.c.  ammonia  (10  per 
cent.)  were  then  added  and  the  mixture  shaken  four  hours  in  a 
shaker.  Thirty-five  c.c.  water  were  now  added,  and  after  shaking 
and  settling  of  the  drug,  100  grammes  (=  10  grammes  drug)  were 
drawn  off  and  shaken  out  with  three  portions  ot  acid  water,  using 
50  c.c,  40  c.c.  and  30  c.c.  successively.  The  alkaloid  was  then 
shaken  out  with  light  chloroform-ether  mixture  and  ammonia,  and 
estimated  alkalimetrically  by  my  general  method,  using  Wagner's 
reagent  as  precipitant. 
(2)  The  same  amount  of  the  same  drug  was  treated  in  exactly  the 
same  manner,  only  leaving  out  the  addition  of  water. 
The  results  obtained  were  as  follows : 
N  Acid  Consumed 
40  Percentage  of 
by  10  Grammes.  Alkaloid. 
(1)  With  water  9    c.c.  0*648 
(2)  Without  water   .  y6  c.c.  0*576 
The  addition  of  water  then  raised  the  results  nearly  1 1  per  cent. 
It  is  for  this  reason  that  in  method  B  the  addition  of  water  was  left 
out  altogether.  The  ethereal  layer,  in  most  cases  in  my  experience, 
separates  out  quite  well,  even  without  the  addition  of  water,  pro- 
vided it  is  set  aside  for  a  sufficient  length  of  time.  In  some  cases  the 
shaking  up  of  the  mixture,  before  drawing  off  the  aliquot  part,  with 
about  2  grammes  of  calcined  magnesia,  will  clear  up  the  superna- 
tant layer  very  quickly. 
