﻿240 
  ANNUAL 
  REPORT 
  SMITHSONIAN 
  INSTITUTION, 
  19 
  2 
  9 
  

  

  increase 
  the 
  surface 
  as 
  much 
  as 
  possible. 
  In 
  order 
  to 
  secure 
  this 
  a 
  

   suspension 
  of 
  pure 
  aluminium 
  powder 
  in 
  water 
  maintained 
  by 
  a 
  

   stream 
  of 
  carbon 
  dioxide, 
  was 
  exposed 
  to 
  ultra-violet 
  light. 
  Increased 
  

   yields 
  of 
  carbohydrates 
  were 
  at 
  once 
  obtained, 
  but 
  it 
  was 
  also 
  found 
  

   that 
  the 
  best 
  yields 
  were 
  obtained 
  when 
  the 
  aluminium 
  powder 
  had 
  

   been 
  allowed 
  to 
  coat 
  itself 
  with 
  hydroxide 
  by 
  remaining 
  in 
  contact 
  

   with 
  the 
  water 
  before 
  the 
  exposure 
  to 
  light. 
  This 
  latter 
  observation 
  

   very 
  materially 
  changed 
  our 
  ideas, 
  since 
  it 
  established 
  the 
  fact 
  that 
  

   the 
  surface 
  phenomenon 
  is 
  of 
  far 
  greater 
  importance 
  than 
  the 
  re- 
  

   ducing 
  action, 
  and 
  indeed 
  raised 
  the 
  question 
  as 
  to 
  vv^hether 
  the 
  latter 
  

   plays 
  any 
  role 
  at 
  all 
  in 
  the 
  process. 
  

  

  In 
  order 
  finally 
  to 
  decide 
  this 
  question 
  an 
  aqueous 
  suspension 
  of 
  

   pure 
  and 
  freshly 
  prepared 
  aluminium 
  hydroxide, 
  maintained 
  by 
  a 
  

   stream 
  of 
  carbon 
  dioxide, 
  was 
  exposed 
  to 
  ultra-violet 
  light. 
  There 
  

   was 
  obtained 
  after 
  filtration 
  and 
  evaporation 
  of 
  the 
  solution 
  a 
  quantity 
  

   of 
  carbohydrates 
  equal 
  in 
  weight 
  to 
  that 
  produced 
  when 
  aluminium 
  

   powder 
  was 
  used. 
  This 
  conclusively 
  proved 
  the 
  fundamental 
  signifi- 
  

   cance 
  of 
  the 
  role 
  played 
  by 
  the 
  surface, 
  and 
  at 
  the 
  same 
  time 
  the 
  

   reducing 
  action 
  was 
  found 
  to 
  be 
  entirely 
  unnecessary. 
  Identical 
  

   results 
  were 
  obtained 
  with 
  other 
  powders, 
  such 
  as 
  aluminium, 
  zinc, 
  

   and 
  magnesium 
  carbonates. 
  

  

  During 
  the 
  course 
  of 
  these 
  experiments 
  it 
  occurred 
  to 
  one 
  of 
  my 
  

   students 
  (Dr. 
  W. 
  E. 
  Stephen) 
  that 
  if 
  a 
  green 
  powder 
  were 
  used 
  in 
  place 
  

   of 
  the 
  white 
  ones 
  the 
  photosynthesis 
  might 
  take 
  place 
  in 
  visible 
  light, 
  

   the 
  green 
  color 
  being 
  suggested 
  by 
  the 
  green 
  color 
  of 
  the 
  plant- 
  

   pigment 
  chlorophyll. 
  This 
  was 
  found 
  actually 
  to 
  be 
  the 
  case, 
  since 
  

   a 
  suspension 
  of 
  nickel 
  carbonate 
  maintained 
  by 
  a 
  stream 
  of 
  carbon 
  

   dioxide 
  in 
  water, 
  on 
  exposure 
  to 
  the 
  light 
  from 
  an 
  ordinp.ry 
  tungsten 
  

   filament 
  lamp, 
  gave 
  a 
  larger 
  yield 
  of 
  carbohydrates 
  than 
  any 
  of 
  the 
  

   white 
  powders 
  in 
  ultra-violet 
  light. 
  We 
  soon 
  found 
  that 
  there 
  was 
  

   no 
  especial 
  virtue 
  in 
  the 
  green 
  color, 
  and 
  that 
  equally 
  good 
  results 
  

   were 
  given 
  by 
  pink 
  cobalt 
  carbonate. 
  

  

  Apart 
  from 
  the 
  interest 
  which 
  accrues 
  from 
  the 
  fact 
  that 
  the 
  

   photosynthesis 
  is 
  thus 
  achieved 
  in 
  a 
  way 
  which 
  shows 
  a 
  real 
  analogy 
  

   with 
  the 
  natural 
  phenomenon, 
  the 
  method 
  with 
  a 
  colored 
  surface 
  

   and 
  visible 
  light 
  has 
  the 
  very 
  material 
  advantage 
  in 
  that 
  the 
  danger 
  

   of 
  photochemical 
  decomposition 
  by 
  ultra-violet 
  light 
  is 
  completely 
  

   eliminated, 
  with 
  the 
  result 
  that 
  the 
  products 
  are 
  obtained 
  in 
  a 
  purer 
  

   state. 
  

  

  From 
  the 
  above 
  description 
  of 
  the 
  direct 
  photosynthesis 
  of 
  carbo- 
  

   hydrates 
  from 
  carbonic 
  acid 
  in 
  the 
  laboratory 
  several 
  points 
  arise 
  

   which 
  require 
  discussion 
  and 
  explanation. 
  In 
  the 
  first 
  place 
  it 
  may 
  

   be 
  stated 
  that 
  the 
  most 
  rigid 
  control 
  experiments 
  which 
  we 
  could 
  

   devise 
  have 
  definitely 
  established 
  the 
  fact 
  that 
  the 
  carbohydrates 
  

   are 
  not 
  due 
  to 
  the 
  presence 
  of 
  impurities. 
  

  

  