ALCOHOL  AS  A  SOURCE  OF  ERROR  IN  THE  TITRATION 
OF  ALKALOIDS  AND  ALKALOID AL  RESIDUES.1 
Alcohol  in  Titration  of  Alkaloih^n^  4Jg 
By  Chas.  Caspari,  Jr. 
Methods  for  the  volumetric  determination  of  alkaloids  in  crude 
drugs  and  galenical  preparations  frequently  include  directions  to 
dissolve  the  varnish-like  residue  (after  the  same  has  been  washed 
with  ether  and  dried  to  constant  weight)  in  alcohol,  with  the  aid  of 
heat  if  necessary,  and  then  to  add  water  until  a  slight  permanent 
turbidity  results.  A  definite  quantity  of  decinormal  acid,  sufficient 
to  insure  a  slight  excess,  having  been  added  to  the  mixture,  the 
excess  is  titrated  with  centinormal  alkali  in  the  presence  of  a  suit- 
able indicator. 
In  the  course  of  some  recent  analytical  work,  the  writer  observed 
that  alcohol  appeared  to  influence  the  color  produced  by  acids  and 
alkalies  with  different  indicators  in  the  titration  of  alkaloidal  resi- 
dues, and  a  series  of  experiments  were,  therefore,  made  to  study 
more  closely  the  nature  of  the  changes  observed,  and  also  to  deter- 
mine, if  possible,  whether  alcohol  really  was  the  disturbing  factor. 
Plain  water,  diluted  alcohol  (a  mixture  of  equal  volumes  of  alco- 
hol and  water),  94-5  per  cent,  alcohol  (commercially  known  as 
cologne  spirit)  and  absolute  alcohol,  were  employed  in  connection 
with  decinormal  sulphuric  acid  and  centinormal  potassium  hydrox- 
ide solution,  as  also  the  following  well-known  indicators  ;  haema- 
toxylin,  cochineal,  Brazil  wood,  methyl  orange  or  tropaeolin  OO, 
lacmoid  and  litmus.  Tap  water  was  found  unfit  for  colorimetric 
work,  as  it  invariably  caused  an  alkaline  reaction  with  the  indi- 
cators, even  after  having  been  well  boiled,  and  pure  distilled  water 
was,  therefore,  employed  instead.  10  c.c.  of  the  respective  liquids 
were  put  into  a  beaker,  together  with  the  indicator,  and  acid  or 
alkali  added  until  the  desired  change  of  color  was  produced. 
The  following  results  are  very  significant  and  well  worthy  of 
attention  : 
Hematoxylin  solution,  1  gm.  to  too  c.c.  alcohol.    Three  drops  were  used  for 
each  experiment. 
10  c.c.  distilled  water;  the  addition  of  1  drop  t£q  KOH  sol.  caused  a  decided 
purple  color. 
1  Presented  at  the  Montreal  meeting  of  the  American  Pharmaceutical  Asso- 
ciation. 
