466  Estimation  of  Glycerin  in  Fats.  {AmsJeXSTm' 
In  a  similar  experiment  the  diluted  solution  was  vigorously  boiled 
down  over  the  naked  flame  to  about  one  half  (the  basin  being,  of 
course,  covered  with  a  dock  glass  to  prevent  loss  by  spurting),  and 
heated  for  two  hours  with  sulphuric  acid  and  bichromate.  Found 
•2961  grms.  or  98.5  per  cent,  of  the  glycerin  taken. 
In  a  precisely  analogous  experiment  10  c.c.  of  strong  alcohol  were 
added  to  the  water,  before  boiling,  over  the  naked  flame.  After  two 
hours  30*33  c.c.  of  chromate  were  reduced,  equal  to  *2973  grms.  or 
98*9  per  cent,  of  glycerin  found. 
The  alcohol  experiment  repeated,  but  the  fluid  concentrated  on  the 
water  bath,  an  amount  of  bichromate  was  reduced  corresponding  to 
107*6  per  cent,  of  glycerin.  An  odor  of  aldehyde  was  perceptible 
during  the  oxidation. 
Deductions. — From  these  test  experiments  the  following  conclusions 
can  be  drawn : 
1.  In  a  fairly  concentrated  solution  glycerin  is  quantitatively  oxi- 
dized by  acid  bichromate ; 
2.  In  a  very  dilute  solution  the  oxidation  is  not  complete  even  after 
many  hours'  heating; 
3.  The.  addition  of  hydrochloric  acid  does  not  materially  help  oxi- 
dation ; 
4.  In  solutions  containing  about  10  per  cent,  of  strong  sulphuric 
acid  the  oxidation  is  complete  after  two  hours,  even  in  exceedingly  di- 
lute solutions  (6  glycerin  per  1000  of  fluid); 
5.  From  such  dilute  solutions,  glycerin  does  not,  as  is  commonly  as- 
sumed, volatilize  on  concentrating  the  fluid,  be  it  on  the  loater-bath  or 
over  the  naked  flame; 
6.  Should  alcohol  be  present  it  is  completely  volatilized  by 
vigorously  boiling  the  fluid  down  to  one  half,  but  not  on  the  water- 
bath. 
The  non-volatility  of  glycerin  from  dilute  solutions  may  further  be 
readily  demonstrated  by  distilling  from  a  large  retort  500  c.c.  of  wa- 
ter, containing  about  *3  grms.  of  glycerin,  catching  the  first  250  c.c.  of 
the  distillate.  This  distillate  does  not,  even  on  heating,  decolorize 
more  than  a  few  drops  of  a  dilute  permanganate  solution,  such  as  is 
used  in  water  analysis. 
Method  for  estimating  glycerin  in  fats. — Saponify  about  3  grms. 
of  the  fat  with  alcoholic  potash ;  do  not  drive  off  all  the  alcohol,  lest 
glycerin  should  volatilize  from  the  concentrated  solution,  but  dilute 
