AmAJp°rLr;imrm"}  Oil  of  Bitter  Almonds.  157 
be  modified,  because  it  is  becoming  possible  to  obtain  benzaldehyde 
produced  by  chemical  synthesis  which  shows  no  presence  of  these 
intermediate  chlorine-bearing  compounds.  All  the  substitutions  of 
genuine  oil  of  bitter  almonds  by  benzaldehyde  occurring  in  the  oils 
here  recorded  have  been  easily  detected  by  the  presence  of  such 
chlorinated  bodies.  It  is  interesting  to  note  in  this  connection  that 
there  was  at  one  time  a  controversy  between  the  chemists  of  Merck 
&  Co.,  and  those  of  Schimmel  &  Co.,  the  former  claiming  to  have 
found  traces  of  chlorine  even  in  genuine  oil  of  bitter  almonds.  This 
fact  was  subsequently  proven  to  be  erroneous  by  Schimmel  &  Co., 
and  they  at  that  time  proved  the  efficacy  of  their  qualitative  test* 
giving  also  processes  for  the  quantitative  estimation  (comparatively 
at  least)  of  chlorine  present  both  in  the  substituted  compounds  and 
the  organic  chlorides  (SchimmeVs  Report,  April,  189 1,  3). 
Copper  Test. — This  is  the  well-known  test  for  the  detection  of 
halogens  in  organic  compounds,  and  its  exceeding  delicacy  makes  it 
very  valuable,  though  it  is  probably  no  more  delicate  than  the  silver 
nitrate  test ;  but  at  least  it  is  not  undesirable  to  have  two  such 
excellent  tests  to  check  each  other,  particularly  when  we  consider 
possible  sources  of  error  in  silver  nitrate  tests,  to  which  we  will 
have  occasion  to  refer  later.  By  this  test  only  five  of  the  twenty- 
two  samples  were  shown  to  be  absolutely  free  from  any  trace  of 
chlorine.  These  are  numbers  I,  3.  8,  9  and  10,  and  hence  may  be 
considered  genuine  oils,  the  remainder  being  in  whole  or  in  part 
artificial  benzaldehyde. 
Silver  Nitrate  Test. — This  test,  which  is  the  only  one  given  in  the 
U.SP.,  1 890,  is  very  reliable,  it  being  both  extremely  delicate  when 
properly  handled  and  capable  of  giving  some  approximate  idea  of 
the  quantity  of  organic  chlorine-containing  bodies  present.  The 
only  difference  between  this  and  the  preceding  test  is  that  while 
inorganic  and  non-volatile  chlorides  will  give  no  indications  by  this 
test,  their  presence  will  be  demonstrated  by  the  copper  test.  The 
U.S.P.  test  is  so  worded  as  to  exclude  any  possibility  of  error  due 
to  the  production  of  silver  cyanide  instead  of  silver  chloride,  and 
the  consequent  misinterpretation  of  the  test,  but  it  fails  completely 
to  take  cognizince  of  another  and  very  much  more  important 
source  of  error,  and  one  which  I  do  not  remember  having  seen  any 
notice  of  in  this  immediate  connection,  viz.:  the  possibility  of  the  filter 
paper  used  containing  chlorides. 
