170 Prof. B. Moore and Mr. T. A. Webster. 



are ions and molecules in the ferric chloride solution of greater molecular 

 complexity than the mono-molecular condition. 



The explanation then of the greater effect with undialysed ferric chloride, 

 as contrasted with colloidal ferric oxide solution, is that the mass of the 

 complex molecule in the ordinary ferric chloride solution is that which can 

 best take up the light vibrations and absorb the energy, whereas in the 

 colloidal ferric oxide solution the solution-aggregate has become too heavy to 

 take up the light-energy and convert it into chemical energy. 



On the other hand, with the uranium salts, the optimum solution-aggregate 

 to act as a transformer for light-energy lies nearer that found in the dialysed 

 solutions than in the ordinary uranium nitrate solutions, and accordingly the 

 former show a higher photo-synthetic activity. 



Silicic acid and its salts show the same kind of effect as the uranium 

 compounds, for dialysed silicic acid solution gives a strong photo-synthetic 

 action, while both sodium silicate solution and undialysed silicic acid solution 

 are inactive. 



There appear, therefore, to be two factors in the production of photo- 

 synthetic activity by a given light source, viz., (1) the specific character of 

 the inorganic catalyst or transformer, (2) the degree of its molecular aggrega- 

 tion in solution. 



In view of criticisms that have been made as to the necessity of inorganic 

 catalysts in the solution, and as to the possibility of the ultra-violet rays 

 producing the synthetic effect when passed into pure water charged with 

 carbon dioxide, as also the view that the formaldehyde obtained might be 

 produced from the minute traces of organic matter in the dialysed solutions 

 and not synthetically from carbon dioxide and water, the following critical 

 series of experiments was carried out. 



1. Water alone, freshly re-distilled, was saturated by a stream of carbon 

 dioxide and then exposed in a quartz test-tube during the whole of a bright 

 summer day on the roof to direct sunlight ; alongside it in a similar quartz 

 test-tube was exposed a 1-per-cent. solution of ferric chloride also saturated 

 with carbon dioxide, and a third test-tube filled with 1-per-cent. ferric 

 chloride saturated with carbon dioxide was kept in a dark cupboard. 



Tested at the end, after distilling away from the iron salts, the distilled 

 water tube and dark ferric chloride tube gave a negative result, while the 

 tube of ferric chloride exposed to sunlight in presence of carbon dioxide 

 gave a bright pink colour with the Schryver's reagent corresponding to about 

 1 in 500,000 of formaldehyde. 



2. A solution of 1-per-cent. ferric chloride was made up with distilled 

 water which had been freshly boiled and was free from carbon dioxide ; this 



