594  Fire-Proofing  Treatment  of  Wood.  (^eMmb^S' 
I  have  before  referred  to  the  way  in  which  sulphate  or  phosphate 
of  ammonia  acts  to  make  wood  fire-resistant,  viz.,  by  rapidly  liberat- 
ing ammonia  gas,  which  has  the  effect  of  checking  the  flames  on 
the  surface  of  the  wood.  The  fiercer  the  flame  which  plays  against 
such  wood  the  more  rapid  the  liberation  and  exhaustion  of  the  pro- 
tecting vapor.  There  is  no  residual  protective  substance  remaining 
in  the  wood,  and  the  carbonization  of  the  fibre  proceeds  apace. 
On  the  other  hand,  so  soon  as  the  sulphate  of  alumina  of  the 
superficial  layer  of  the  wood  impregnated  with  this  chemical  is 
decomposed  by  the  heat  of  a  flame,  a  deposit  of  alumina  is  formed, 
the  non-conducting  properties  of  which  make  it  a  barrier  against 
the  propagation  of  the  carbonizing  effect  and  protect  the  interior 
in  a  very  notable  degree.  An  actual  experiment,  one  of  a  large 
number  which  1  carried  out  jointly  with  the  inventor,  will  illustrate 
this.  If  a  piece  of  wood  be  saturated  with  a  solution  of  sulphate  of 
alumina  of  300  B.  strength  to  a  depth  of  not  more  than  ^  of  an 
inch  from  the  surface  and  the  point  of  the  inner  blue  cone  of  a 
strong  Bunsen  flame  be  made  to  impinge  upon  it  and  kept  in  such  a 
position,  a  boring  effect  takes  place,  while  an  abundant  separation 
of  alumina  will  be  observed.  The  average  resistance  of  a  piece  of 
i-inch  white  pine  so  treated  to  the  complete  boring  result,  with  final 
penetration  of  the  flame  to  the  other  side,  wrll  be  over  three  hours. 
If  a  similar  piece  of  I -inch  white  pine  be  "  heart-saturated  "  with 
ten  times  the  quantity  of  sulphate  of  ammonia  and  the  same  Bunsen 
flame  be  applied  under  exactly  similar  conditions,  the  average  resist- 
ance to  complete  penetration  will  not  be  over  seventy  minutes. 
These  results  have  been  obtained  repeatedly  and  in  instances  the 
disproportion  was  much  greater. 
Some  very  interesting  observations  have  been  made  on  the  physi- 
cal changes  which  the  fireproofing  material  undergoes  on  the  con- 
tinued application  of  heat.  As  a  result  of  repeated  measurements, 
it  is  found  that  the  residual  alumina  occupies  a  space  from  two  and 
a  half  to  three  times  as  great  as  the  dried  salt  from  which  it  is 
formed.  Hence  in  forming  it  apparently  expands  to  fill  out  the  air 
spaces  and  intercellular  spaces  of  the  wood  very  fully.  This  results 
in  the  formation  of  a  very  compact  non-conducting  barrier  which 
interposes  itself  to  the  action  of  the  flame  and  protects  the  layers 
of  woody  tissue  upon  which  it  is  formed.  The  protection  is  there- 
fore a  real  and  much  more  lasting  one  than  that  which  could  come 
