Amjufyri92iarm' }        Glycyrrhizin  in  Licorice  Root.  471 
2?.  Schmidt  (Haffner),  (191 1). 
[Translator's  Note— Schmidt  uses  Haffner's  mixture  of  alco- 
hol-sulphuric acid  in  treating  the  original  powdered  licorice  extract, 
but  instead  of  going  through  the. acetone  treatment,  and  obtaining 
the  barium  salt,  he  weighs  the  precipitated  glycyrrhizin  directly  on 
a  tared  filter  after  washing  with  dilute  sulphuric  and  then  a  little 
water. 
Linz  comments  on  the  absence  of  exact  instructions  as  to  quan- 
tities of  sulphuric  acid  with  which  to  precipitate  and  to  wash.  He 
obtained  9.1  and  94  per  cent,  glycyrrhizin,  and  appears  to  have  little 
objection  to  the  method.  (P.  A.  H.)] 
Conclusions  From  My  Researches. 
As  the  result  of  my  researches,  I  have  made  a  table  in  Appendix 
C  which  summarizes  the  methods,  and  which  is  intended  to  serve  in 
deciding  the  question  as  to  which  is  the  best  method. 
A  large  number  of  the  methods  are  summarily  rejected.  I  so 
classify  all  those  determinations,  which,  through  false  analytical 
procedure,  give  results  which  are  certainly  wrong — too  high  or  too 
low.  Then  I  also  reject  such  methods  which,  through  inexact  or 
erroneous  directions,  give  results  which  are  unreliable,  and  not  suit- 
able for  direct  comparison. 
In  the  first  class  belong  the  method  of  Capin,  Erikson,  Trubeck, 
the  Dutch  and  French  Pharmacopoeias. 
In  the  second  class  belong  Rump,  Helfenberg,  Kremel,  Diehl, 
Guignard,  Py,  Anselmino-Gilg,  Stoeder,  Schroder,  Miintzer,  Kin- 
zey,  Durier,  Schmidt-HafTner.  In  this  second  class  I  have  placed 
all  those  methods  which  do  not  and  cannot  give  directly  reliable  re- 
sults when  carried  out  exactly  as  the  directions  call  for.  In  so  doing, 
however,  the  analytical  procedure  on  which  the  method  is  based  is 
not  in  itself  designated  as  impracticable,  but  only  the  execution  of 
the  method,  so  that  the  particular  methods  in  this  class  yield  prac- 
tical results,  after  certain  modifications. 
Less  practical  are  the  methods  of  Guignard,  Gadais  II,  Telle 
and  Cederberg,  for  reasons  which  I  have  given  in  discussing  the  indi- 
vidual methods. 
Practical  results  are  given  by  the  three  almost  identical  methods 
of  Parry,  Evans'  Sons,  Lesher  &  Webb,  and  Houseman,  and  per- 
haps also  Gadais  I. 
