TO THE ORIGIN OF LIFE 43 



alongside, and given the same exposure and general conditions, we 

 have been able to show that the colloidal condition is much more 

 active in this photo- chemical reaction. In this comparative ex- 

 periment the crystalloidal uranium nitrate was precipitated while 

 no precipitation whatever occurred in the colloidal uranic hydroxide. 

 Passing on, we have shown that similar photo- synthesis of organic 

 from inorganic matter easily occurs with colloidal ferric hydroxide, 

 and by the use of the " Uviol " mercury lamp we are enabled to 

 experiment at all times and obtain synthetic results readily in a 

 few hours, a consideration of some importance in working in a 

 country where sunlight is so variable, and so often unavailable for 

 days. 



The experiments with uranic hydroxide and ferric hydroxide 

 were made concurrently in time, but for convenience we shall 

 describe first the uranium experiments and then those with tne 

 ferric hydroxide. 



PHOTO-SYNTHESIS BY COLLOIDAL URANIC HYDROXIDE. 



Method of Preparation of the Colloidal Solution. A strong solu- 

 tion of uranium nitrate (approximately 10 per cent.) is taken and 

 treated in the cold by adding a saturated solution of ammonium 

 carbonate until the precipitate just ceases to redissolve. The solu- 

 tion is then filtered and dialysed in a tube of parchment paper for 

 several days against running water. The greater part of the 

 uranium is still in the crystalloidal state, and dialyses away at this 

 state. Great care is required in the first step to get the proper 

 relative amount of ammonium carbonate; also it is as well not to 

 wait for the removal of the last traces of crystalloidal uranium, 

 but to take the solution when it still gives a faint reaction with 

 potassium ferrocyanide. The amount so left is, however, very 

 minimal compared with the crystalloidal uranium solutions as used 

 in the experiments recorded below, and nearly all the uranium at 

 the end is present as colloidal uranic hydroxide. 



This method is practically the same as that described by Graham 

 for the preparation of colloidal ferric hydroxide and as used in the 

 preparation of our colloidal iron solutions. 



The amount of uranium in this solution was determined by 

 evaporating a measured volume to dryness and weighing the 

 residue, and was found to be 0478 per cent, of Ur 2 a . 



