AmAP?Ci87h7!rm'}    Detection  of  Castor  Oil  in  Copaiba.  183 
platinum  dish  till  dry,  and  then  weighed,  and  the  weight  of  the  wax 
used  deducted.  The  beaker,  filter  and  rod,  etc.,  used  are,  if  at  all 
dirty,  dried,  extracted  with  ether,  and  the  residue  left,  after  evapora- 
tion weighed  and  added  to  the  total. 
The  calculation  is  then  performed  as  follows  : 
1.  To  the  weight  in  grams  found  add  "io  for  loss  of  oleic  acid  in 
solvent,  and  then  say  as 
95  :  100  ::  total  oleic  acid. 
2.  Calculate  to  per  cent,  from  the  quantity  taken,  and  from  the 
total  per  centage  deduct  six  per  cent,  for  possible  altered  resin  in  the 
balsam. 
Out  of  the  whole  number  of  samples  I  have  done,  I  have  selected 
the  following  twelve  as  being  fair  representations  of  the  degree  of 
accuracy  obtainable  by  the  process.  The  error,  owing  to  the  correc- 
tion, of  course,  increases  with  the  amount  of  oil  present,  but  it  is 
always  an  error  in  the  direction  of  under-estimation,  which  is  the  great 
point  for  public  analysts. 
Nature  of  Sample. 
Para  (pale)  ..... 
Para  (pale)  ..... 
Old  Para  (dark)  
Old  Para  (dark).  .... 
Carthage  (medium) 
Carthage  (medium)  .... 
Maranham  (pale)  .... 
Maramham  (pale)  ...... 
Old  Maranham  (darkish,  very  little  odor) 
Old  Maranham  (darkish,  ver)'  little  odor)  , 
Para  (fine  pale)  ..... 
Para  (fine  pale)  . 
In  conclusion,  I  may  say  that  the  process,  although  it  looks  formida- 
ble, is  in  practice  very  simple,  and  for  all  ordinary  purposes,  if  the 
beaker  be  well  scraped  out,  the  weight  of  the  main  cake  may  be  taken 
as  sufficient  to  give  an  analysis  true  within  3  per  cent,  below  the  real 
amount,  which  is  accurate  enough  for  public  purposes,  and  saves  time 
and  the  expense  of  the  extra  ether.  Unless  oil  actually  floats  and 
remains,  on  cooling,  in  fluid  drops,  after  adding  the  hydrochloric  acid,  the 
sample  may  be  passed  as  good. 
When  working  on  three  to  four  grams,  with  an  admixture  of  not 
over  25  per  cent.,  the  errors  due  to  loss  of  oleic  acid  and  insoluble 
Calculated. 
Found. 
Pure 
No  oil  drops. 
23-60  percent,  castor 
23-50 
Pure 
No  oil  drops . 
51*0  per  cent. 
castor 
5o-o  per  cent. 
Pure 
No  oil  drops. 
21  -5  per  cent. 
castor 
21'20 
Pure 
No  oil  drops. 
26*5  per  cent. 
castor 
26.27 
Pure 
No  oil  drops. 
47-3  per  cent. 
castor 
46-4 
Pure 
No  oil  drops. 
2i*4  per  cent. 
lard  oil 
20  9 
