266 
DRYING  OP  OILS  BY  METALLIC  SALTS. 
In  the  manufacture  of  fatty  drying  oils,  the  oxides  taken  up 
by  them  are  not  completely  reduced,  by  which,  as  we  have  con- 
vinced ourselves,  carbonic  acid  is  generated :  the  reduced  oxide 
is  thereby  converted  into  a  body,  which  acts  on  the  oil  like  a  fer- 
ment ;  the  conclusion  is,  therefore,  that  boiled  linseed  oil  has  in 
no  way  drying  properties  when  it  contains  no  dissolved  oxide. 
We  found  that  the  bodies  which  possess  this  inciting  power  in 
the  greatest  degree,  are  most  protoxides  of  the  metals  Mn,  Zn, 
Fe,  Sn,  Co,  Cd,  and  among  these  the  protoxides  of  cobalt  and 
manganese  afforded  us  the  most  satisfactory  results.  In  some 
^ases  the  protoxide  of  iron  behaved  in  a  similar  manner,  how- 
ever less  powerfully. 
To  find  a  ferment  or  siccative  which  acts  rapidly  without  in- 
jury on  the  drying  oils,  we  must  consequently  seek  for  com- 
pounds of  the  above-mentioned  oxides,  which  are  readily  to  be 
prepared  on  a  large  scale,  and  which  allow  these  oxides  to  exer- 
cise their  properties  on  the  oil.  This  is  not  the  case  with  the 
above-mentioned  oxides  when  they  are  difficult  to  prepare,  and 
cannot  be  preserved  in  contact  with  the  air. 
"We  directed  our  attention,  therefore,  in  «the  first  place  to  the 
inorganic  and  organic  compounds  of  cobalt  and  manganese.  We 
found  that  carbonic,  phosphoric,  sulphuric,  nitric,  and  muriatic 
acids,  as  well  as  most  organic  acids,  retained  their  oxides  too 
firmly  and  almost  destroyed  their  action.  The  salts  of  these 
acids  in  a  basic  condition  have  a  distinct  influence. 
But  of  all  inorganic  acids,  boracic  acid  is  that  which  afforded 
the  most  satisfactory  results  in  combination  with  the  protoxide 
of  cobalt  and  manganese.  The  proportion  of  borate  of  manga- 
nese, which  suffices  to  induce  fermentation  in  a  drying  oil,  is 
from  1  to  1 J  thousandth  part  of  the  weight  of  the  oil. 
We  may  remark  that  the  borate  of  manganese  which  we  em- 
ploy, and  with  the  study  of  which  we  are  now  engaged,  is  not 
an  anhydrous  salt,  but  contains  25  per  cent,  of  water.  It  seems 
to  us  to  act  in  the  following  manner : — A  portion  of  the  pro- 
toxide is  separated  under  the  influence  of  light  and  heat,  it  ab- 
sorbs the  oxygen  of  the  air  to  form  the  proto-peroxide,  and  then 
it  is  noticed  that  the  oil  begins  to  become  adhesive.  It  is  also 
remarkable,  that  the  film  at  this  moment  acquires  a  color  which 
disappears  when  the  film  is  dry.    If  1  to  2  per  cent,  of  borate  of 
