^ptJe0mrbe?,hS'}     Alcohol  in  Titration  of  Alkaloids.  475 
b.  10  c.c.  distilled  water  required  2  drops  T5T  KOH  sol.  to  produce  a  decided 
purplish  blue  color. 
10  c.c.  diluted  alcohol  required  0*2  c.c.  x£o  KOH  sol.  to  produce  the  same  color 
as  in  a. 
10  c.c.  diluted  alcohol  required  0*65  c.c.  x£o  KOH  sol.  to  produce  the  same 
color  as  in  b.  The  color  was  again  destroyed  by  addition  of  a  few 
drops  of  alcohol. 
10  c.c.  alcohol  required  no  c.c.  x^  KOH  sol.  to  produce  the  same  color  as  in  b. 
10  c.c.  absolute  alcohol  produced  the  same  color  as  obtained  in  b  with  the 
indicator  alone. 
\ 
Tropczolin  OO  or  methyl  orange  solution,  1  gm.  to  500  c.c.  50  per  cent,  alco- 
hol.   Two  drops  were  used  for  each  experiment. 
a.  10  c.c.  distilled  water,  upon  addition  of  1  drop  r^  H2S04,  gave  the  charac- 
teristic pink  color,  showing  an  acid  reaction. 
b.  10  c.c.  distilled  water,  with  o'i  c.c.  ^  H2S04,  gave  a  decided  crimson  color, 
showing  a  strong  acid  reaction. 
c.  10  c.c.  diluted  alcohol  required  no  c.c.  ^  H2S04  to  produce  the  same  color 
as  in  b. 
10  c.c.  alcohol,  with  3-5  c.c.  -~  H2S04,  failed  to  produce  the  same  color  as  in  b  ; 
a  deep  orange-red  color  was  produced,  which  gradually,  on  further  addition 
of  1  '25  c.c.      H2S04,  changed  t'o  crimson. 
10  c.c.  absolute  alcohol  failed  to  produce  a  crimson  color  with  4  75  c.c. 
to  H2S04. 
d.  10  c.c.  distilled  water,  treated  as  under  b,  required  0*97  c.c.        KOH  sol.  to 
produce  a  strong  yellow  color  indicating  alkalinity. 
10  c.c.  diluted  alcohol,  treated  as  under  c,  required  only  io-20  c.c.  T^o  KOH 
sol.  to  produce  the  same  color  as  in  d. 
From  the  foregoing  reactions  it  is  very  evident  that  alcohol  and 
absolute  alcohol,  as  available  in  the  market,  exercise  a  decided 
influence  on  color  indicators,  and  may  be  the  fruitful  source  of  error 
in  volumetric  work.  Strange  to  say,  while  alcohol  appears  to  play 
the  part  of  an  acid  toward  hematoxylin,  cochineal,  Brazil  wood, 
lacmoid  and  litmus,  by  requiring  an  increased  quantity  of  alkali  to 
produce  the  characteristic  alkaline  color  reaction,  it  behaves  quite  dif- 
ferently towards  methyl  orange  or  tropaeolin  OO.  In  the  latter  case 
alcohol  seems  to  lend  to  the  indicator  a  strong  alkaline  reaction, 
requiring  a  phenomenal  amount  of  decinormal  acid  to  produce  the 
characteristic  acid  color.  The  fact  that  absolute  alcohol  appears 
alkaline  towards  all  of  the  above  indicators  is  remarkable,  and,  while 
no  further  examination  of  the  article  was  undertaken,  it  is  but  fair 
