50 SYNTHESIS BY SUNLIGHT IN RELATIONSHIP 



per cent, of colloid, was placed in a wide 1000 c.c. measuring cylinder. 

 In this glass cylinder was immersed a wide round-bottomed tube of 

 " Uviol " glass of somewhat smaller diameter to serve as a protector 

 between the heated lamp to be placed in the interior and the sur- 

 rounding fluid between the " Uviol " tube and the outer cylinder of 

 glass. The mercury lamp was then set going, and inserted in the 

 centre of the " Uviol " glass protecting tube. The mercury lamp 

 tube, also of " Uviol " glass, has a diameter of 3 cm. approximately, 

 the protecting tube a diameter of 54 cm. externally, and the outer 

 glass cylinder an internal diameter of about 6-2 cm. The light 

 generated at a distance of about 2 cm., after passing two thicknesses 

 of " Uviol " glass and a layer of air of somewhat less than a centi- 

 metre, passed through a layer of the colloidal solution about 4 to 

 5 mm. thick and then reached the outer glass vessel. On account 

 of the rounded bottom of the " Uviol " protecting tube there was 

 a certain considerable volume of the solution at the bottom not so 

 well illuminated. The total volume of colloidal solution at the 

 bottom and in the annular space between the two tubes was 

 approximately 300 c.c. 



A slow, steady stream of washed carbon dioxide was passed 

 through the colloidal solution. 



After about half an hour's running the colloidal solution be- 

 came too warm and began to coagulate, so the current was switched 

 off and the lamp allowed to cool. The whole was then immersed 

 in a wide square- sided glass jar of about 17 litres capacity through 

 which a stream of cold water was continuously circulated. The 

 lamp was now relit and 'kept going for a period of 2 hours, making 

 2 1 hours of illumination in all. In this first experiment with the 

 " Uviol " apparatus, the colloidal solution had coagulated at the 

 end, but this did not happen in later experiments when the heat 

 was better regulated. 



The colloidal solution was removed and tested for formaldehyde 

 in two ways. One portion was distilled as in the previously recorded 

 experiments, and the distillate gave a most marked positive reaction 

 with SchifE's reagent a deep pink coloration. A second part was 

 simply filtered off from the coagulated iron precipitate, and at once 

 tested. It also gave a marked positive reaction both with Schiffs 

 and Mulliken's tests. 



Experiment F. An experiment of the same type as the preceding 

 one. A colloidal solution of 1 in 20 of the same stock solution was 



