A^Jgus'tj89a'm'}    Balsam  Copaiba  vs.  Gurjun  Balsam.  397 
The  conclusions  from  the  above  and  other  data  may  be  summar- 
ized as  follows : 
(1)  The  color  and  the  fluorescence  or  non-fluorescence  are  of  no 
value. 
(2)  The  wide  range  of  the  specific  gravity  of  balsam  copaiba 
makes  that  factor  practically  valueless. 
(3)  The  solubility  or  the  insolubility  of  balsam  copaiba  are 
uncertain  factors  and  cannot  be  relied  on.  One  sample  of  pure 
substance  is  soluble  in  absolute  alcohol,  another  is  not.  One  sam- 
ple of  benzine  will  indicate  the  presence  of  gurjun  balsam,  another 
will  not ;  hence,  is  valueless  as  a  specific  test. 
(4)  The  specific  tests  for  turpentine,  fixed  oils  and  paraffin  oils 
are  reliable. 
(5)  The  ammonia  tests  are  misleading  and  are  consequently  unre- 
liable. 
(6)  The  acid  number  cannot  be  relied  on.  At  first  the  author 
anticipated  valuable  results  from  this  source,  but  was  utterly  dis- 
appointed. 
(7)  Hager's  test  is  not  reliable,  especially  with  less  than  25  per 
cent,  of  gurjun  balsam  present. 
(8)  The  carbon  disulphide  test  is  fairly  reliable  with  a  moderately 
large  per  cent,  of  gurjun  balsam.  Some  varieties  of  balsam  copaiba 
turn  reddish-brown  with  this  test,  and  it  sometimes  becomes  very 
difficult  to  discriminate. 
(9)  The  glacial  acetic  acid  (99-5  per  cent.)  is  perfectly  reliable, 
even  when  not  more  than  5  per  cent,  is  present,  at  least  so  far  as  the 
above  results  indicate. 
The  writer  has  found  that  gurjun  balsam  is  even  more  variable  in 
composition  than  balsam  copaiba,  consequently  an  absolute  decision 
is  reserved  until  a  thorough  test  has  been  made  with  a  fair  number 
of  samples.  This  will  be  done  as  soon  as  sufficient  material  can  be 
collected. 
The  U.  S.  P.  instructs  us  to  detect  gurjun,  balsam  in  balsam 
copaiba  by  heating  a  sample  to  1300  C,  when  it  will  gelatinize  if 
the  former  is  present.  Some  samples  may  gelatinize,  but  a  trial 
proved  that  the  material  employed  in  these  experiments  increased 
only  slightly  in  viscosity,  not  even  assuming  a  semi-gelatinous  state. 
The  test  is  unreliable  in  toto. 
Unless  otherwise  specified,  the  dealer  will  always  supply  a  non- 
solidifiable  product  which  will  comply  with  every  test  in  the  U.  S.  P., 
