288  'XTRTBUTIONS    TO   ECONOMIC    GEOLOGY,  1906,  PART    I, 
I  come  now  to  a  very  interesting  phase  of  my  subject,  although  I  \ 
have  not  studied  it  to  any  extent — namely,  the  bleaching  action  of 
light  on  oils.     A  study  of  the  tables  reveals  an  exceedingly  significant 
fact.     The  after  bleaching  (or  bleaching  due  to  the  action  of  light) 
appears  to  be  much  greater  in  the  more  efficiently  bleached  samples 
than  in  those  which  were  only  slightly  bleached.      I  say  "appears  to 
be,"  because  my  color  scale,  although  comparatively  correct,  is  not 
absolutely  so.     In  other  words,  the  absolute  percentage  of  coloring  j 
matter  in  the  samples  will  not  vary  in  the  same  amount  from  number   j 
to  number.      However,  even  if  we  allow  for  this,  it  seems  almost  cer-  I 
tain  that  this  after  bleaching  can  not  he  altogether  due  to  the  chemical 
action  of  I  he  light  on  the  coloring  substance,  for  in  that  case  all  sam- 
ples should  bleach  a  uniform  amount.      It  seems  necessary,  therefore, 
to  look  elsewhere  for  the  cause  oi'  this  action,  and  as  yet  I  have  been 
unable  to  devise  a  satisfactory  hypothesis. 
It  is  evident  that  the  colloidal  theory  is  most  susceptible  of  proof  by 
paeans  of  microscopic  work,  but  for  lack  of  an  instrument  of  sufficiently 
high  power  I  have  been  able  to  do  but  little  in  this  direction.  Such 
results  as  were  obtained,  however,  tend  to  confirm  this  theory. 
Fuller's  earth,  dry  (Fairbank  No.  2),  appears  to  be  made  up  of  a 
mass  of  small  amorphous  mossy  particles  that  look  something  like 
lumps  of  half-dried  alumina  and  seem  to  contain  many  tiny  air  bubbles. 
By  transmitted  light  many  small  plates  are  seen.  Some  grains  that 
arc  u<>t  plates  are  also  visible,  and  a  very  lew  grains  of  a  color  varying 
Prom  deep  red  to  reddish  yellow.  Some  particles  have  the  look  of 
small  rings  or  groups  of  rings.  I  believe  this  to  be  due  to  the  presence 
of  transparenl  crystalline  plates  which  have  commenced  to  alter  to  an 
opaque  substance  around  the  edge.  Fuller's  earth  treated  with  sul- 
phuric acid  also  shows  these  rings.  Kaolin  shows  extremely  line  par- 
ticles that  are  not  mossy  looking,  and  although  these  particles  are  so 
fine  that  their  form  can  not  be  distinguished,  yet  their  sharpness  of 
outline  and  aoncoherence  point  toward  the  crystalline  structure. 
Limestone  clay  shows  large  quartz  grains,  mossy  lumps,  like  ferric 
hydrate,  and  large  grains  that  are  transparenl  in  places,  but  are 
apparently  covered  with  a  brown  mossy  substance,  like  ferric  hydrate. 
Most  of  the  other  materials  tested  were  also  examined  under  the 
microscope,  hut  nothing  of  any  great  significance  was  noted. 
PRACTICAL    APPLICATION    OF    COLLOIDAL   THEORY. 
If,  as  J  anticipate,  further  investigation  establishes  the4  fact  that 
there  is  a  well-defined  relationship  between  the  rational  composition 
of  fuller's  earth  and  the  class  of  coloring  matters  which  it  removes 
most  efficiently  (see  pp.  268,  284),  there  will  evidently  be  afforded  to 
chemists  a  means  of  predicting  the  value  of  fuller's  earth  and  the  class 
of  work  for  which  it  is  best  adapted,  on  the  basis  of  analysis.     A  little 
