450  Folyhydrio  Alcohols  and  Borax.  {^""sep^im'*™* 
treatment  with  a  small  quantity  of  water  it  yielded  a  mass  of  crystals 
which  were  recognized  by  the  usual  tests  as  boric  acid.  The  original 
substance  imparted  a  vivid  green  tinge  to  the  flame,  was  not  apprecia- 
bly acid  to  test  paper,  but  on  moistening  the  paper  with  water  became 
powerfully  so.  The  acidity  of  the  aqueous  solution  was  not  affected 
by  ebullition.  In  another  experiment  a  mixture  of  anhydrous  borax 
and  anhydrous  glycerol  was  heated  to  120°  C,  the  borax  being  in 
excess ;  the  mass  upon  extraction  with  ether  yielded  the  body  just 
described.  It  was  in  like  manner  decomposed  by  water,  and  glycerol 
was  found  in  the  aqueous  solution  by  mixing  with  lime,  drying  and 
exhausting  with  ether.  Absolute  alcohol  also  extracted  this  substance 
without  decomposition  from  the  original  mass,  but  in  this  case  it  was 
associated  with  sodium  borate.  The  residue,  after  extraction  with 
ether,  was  found  to  contain  a  basic  borate,  probably  sodium  metaborate. 
The  above  experiments  are  conclusive  in  showing  that  the  primary 
action  of  glycerol  upon  sodium  pyroborate  under  the  above  conditions 
results  in  the  formation  of  a  compound  which  is  evidently  a  boric 
ether  of  glycerol,  that  is,  glycerol  in  which  some  of  the  hydroxyl  has 
been  displaced  with  the  formation  of  water  by  the  boric  radical.  This 
ether  may  be  termed  glycerol  boi^in.  It  is  decomposed  by  water,  form- 
ing boric  acid  and  regenerating  glycerol ;  hence  the  acidity  of  aqueous 
mixtures  of  glycerol  and  borax.  The  glycerol  borin  produced  in  the 
manner  above  described  agrees  in  its  main  properties  with  the  boric 
ether  of  glycerol  isolated  by  Schiff  and  Becchi  {Compt  Rend.^  Ixii, 
397)  from  the  action  of  heat  upon  a  mixture  of  glycerol  and  boric 
acid,  which  was  represented  by  the  formula,  C3II5BO3.  Glycerol, 
then,  acts  upon  sodium  pyroborate,  combining  with  half  the  boric  an- 
hydride which  the  salt  contains  to  form  glycerol  borin,  sodium  meta- 
borate remaining. 
Action  of  Mannitol  upon  Borax, — The  experiments  of  Klein  upon 
this  subject,  which  have  been  previously  described,  are  open  to  the 
objection  that  water  was  employed  in  order  to  obtain  the  final  products. 
My  previous  experiments  with  glycerol  had  shown  that  water  annihi- 
lates the  glycerol  borin  which  results  from  the  action  of  glycerol  upon 
borax,  and  therefore  had  a  similar  ether  been  produced  in  the  present 
instance  it  would  have  been  ex  hypothesi  decomposed  by  the  water  era- 
ployed  in  Klein^s  experiments.  For  this  reason,  my  experiments  in 
the  first  place  were  made  in  such  a  manner  as  to  exclude  the  use  of 
water.    Anhydrous  mannitol  and  anhydrous  borax,  the  former  being 
