A^SusiPi89asm'}     Caustic  Soda  Sold  as  Caustic  Potash.  393 
Soluble  impurities,  present  only  in  minute  quantities  in  some  of 
the  samples,  as  alumina  and  lime,  are  not  taken  account  of  in  the 
table. 
This  is  another  instance  which  shows  that  the  pharmacist  should 
not  rely  merely  on  the  appearance  of  his  supplies  of  chemicals  for 
their  identity  and  quality,  and  again  emphasizes  the  necessity  of 
applying  chemical  tests,  not  only  to  detect  intentional  substitutions 
and  adulterations,  but  also  to  discover  such  inadvertencies  as  may 
occur  now  and  then,  even  in  the  most  carefully  conducted  manu- 
facturing and  wholesale  establishments.  In  most  cases  this  can  be 
done  by  very  simple  means. 
Potassium  and  sodium  compounds  can  be  readily  distinguished 
by  means  of  the  familiar  flame-test.  This  requires  simply  a  few 
inches  of  platinum  wire,  for  which  clean  iron  wire  may  be  sub- 
stituted, if  necessary,  and  a  non-luminous  gas  flame.  A  fragment 
of  a  pure  potassium  compound,  taken  upon  a  loop  of  platinum 
wire  and  held  in  a  non-luminous  flame,  colors  the  latter  a  pure 
violet,  without  a  tinge  of  yellow,  which  would  indicate  the  presence 
of  potassium.  Sodium  compounds  color  the  flame  intensely  yellow. 
Crude  potashes  always  contain  small  quantities  of  soda,  but  when  it 
is  not  sufficient  to  cover  entirely  the  violet  of  the  potassium  flame,  it 
is  not  objectionable  for  purposes  that  do  not  require  purified  potash. 
When  no  violet  is  visible  in  the  flame,  the  sample  may  still  consist 
largely  of  potash  or  it  may  be  principally  soda,  and,  when  viewed 
through  a  plate  of  blue  glass,  it  can  be  seen  whether  potash  is  pres- 
ent in  small  or  large  quantity,  by  the  intensity  of  color  visible  through 
the  glass,  through  which  the  yellow  does  not  penetrate.  It  must  be 
kept  in  mind,  however,  that  the  intensity  of  the  potassium  flame 
varies,  not  only  with  the  proportion  of  potash  in  the  sample,  but  also 
with  the  quantity  taken  for  the  test  and  the  heat  of  the  flame. 
Accurate  quantitative  estimations  of  potassium  and  sodium,  in 
presence  of  each  other,  can  be  made  only  by  methods  requiring  con- 
siderable time  and  attention,  but  sufficiently  exact  results  for  ordi- 
nary purposes  may  be  obtained  by  methods  that  are  at  the  command 
of  the  practicing  pharmacist.  The  one  based  on  the  insolubility  of 
potassium  bitartrate  in  mixtures  of  alcohol  and  water  is,  perhaps,  the 
most  convenient  and  satisfactory,  when  great  precision  is  not  re- 
quired. This  is  the  method  directed  by  the  U.S. P.  of  1890  for  the 
determination  of  soda  in  potash.    The  sample,  dissolved  in  a  small 
