I50  Obligations  and  Methods  of  Research,    j^"*-  ^Sllch^^To': 
had  devised  a  new  method  of  obtaining  photographs,  the  use  of  a 
gelatin  film  containing  mercurous  iodide.  I  deemed  it  best,  however, 
to  make  some  search  in  the  matter  and  soon  found  the  process  had 
been  published  eleven  years  before  I  was  born.  It  is  still  more  dis- 
appointing to  find  out  that  some  ancient  Greek  anticipated  one  in 
some  discovery. 
Yet  an  important  observation  must  be  made  here,  namely,  that 
while  the  literature  of  science  contains  a  very  large  mass  of  informa- 
tion, it  must  not  be  accepted  without  question.  Experience  shows 
that  even  the  most  capable  of  investigators  may  be  seriously  mis- 
led. This  is  due  to  several  causes.  Preconceived  notions  often 
blind  ablest  men  to  the  real  facts  of  the  case.  The  mind  too  often 
perceives  what  it  wants  to  perceive.  Even  in  the  material  environ- 
ment of  the  analytic  laboratory,  feeling  and  prejudice  are  not  wholly 
lacking.  A  more  frequent  cause  of  error  is,  however,  the  imper- 
fections of  methods  of  research,  and  the  lack  of  thoroughness  of 
information.  Much  as  one  may  be  astonished  in  examining  the 
literature  of  a  given  subject,  at  the  anticipations  in  results,  one  is 
often  equally  astonished  at  how  much  is  taken  for  granted,  and  how 
such  assumptions  form  the  basis  of  further  investigations.  A 
striking  instance  of  this  occurred  in  the  experience  of  my  co-worker, 
Dr.  William  Beam,  and  myself.  We  had  devised  a  process  of  analy- 
sis which  depended  on  the  use  of  a  strong  sodium  hydroxide  in  gly- 
cerol, with  operation  at  a  moderately  high  temperature.  It  was  vital 
to  the  practical  value  of  the  process  that  no  decided  action  should 
take  place  between  the  alkali  and  the  solvent,  especially  that  no 
substances  of  low  volatility  should  be  formed.  Now  Watt's  Dic- 
tionary, a  well-known  and  carefully  compiled  work,  stated  definitely 
that  when  glycerol  is  heated  with  strong  alkali,  salts  of  some  of  the 
lower  fatty  acids  are  formed.  If  this  was  true,  the  process  was 
worthless.  Careful  experiment  showed  that  it  was  not  true.  A 
good  quality  of  glycerol  heated  well  above  the  boiling  point  of  water 
with  strong  sodium  hydroxide  solution  gave  no  appreciable  traces 
of  any  of  these  acids.  The  explanation  of  the  error  is  probably 
that  the  original  investigator  used,  unknowingly,  an  impure  glycerol. 
The  observation  dated  from  an  early  period  when  the  methods  of 
preparing  and  purifying  glycerol  were  but  imperfectly  known - 
Probably  the  application  of  the  substance  to  the  manufacture  of 
high  explosives,  in  which  an  exceptionally  pure  article  is  needed, 
led  mainly  to  the  improvement  in  the  quality  of  the  commercial 
