90 ACTION OF LIGHT BAYS 



suppose that there should be representatives of intermediate con- 

 densation products lying on the path between formaldehyde and 

 the hexoses. Pribram and Franke, in the papers above referred to, 

 believed they were able to identify glycolaldehyde arising from the 

 condensation of two molecules of formaldehyde, and Inghilleri 

 claims to have isolated the racemic sorbose by means of its osazone. 



It was evident to us that we were dealing with a complex mixture, 

 and although there is a quite distinct effect with phenyl-hydrazine 

 in acetic acid solution, quite different from that of formaldehyde, 

 and giving abundance of coloured precipitate, we have not yet 

 succeeded in isolating a crystalline osazone, though we have on 

 several occasions obtained crystals mixed with amorphous smears. 



In order to prepare large quantities of material, four quartz 

 flasks, each of about 300 c.c. capacity, were completely filled with a 

 4 per cent, solution of formaldehyde, and exposed at a distance of 

 about 4 inches from the lamp for several hours daily for a week. 

 The result was disappointing, on account of the slowness of action. 

 The effect appears, like many effects of light, even in clear solutions, 

 to be all concentrated within a comparatively thin layer lying next 

 the incident surface. Accordingly, as in these larger vessels the 

 volume increases much more rapidly than the exposed surface, the 

 concentration of reducing substance in the solution progresses at a 

 correspondingly slow rate. At the end of the week there was a fair 

 amount of reduction in all four flasks, but not more than would be 

 obtained in a small quartz test-tube with a single day's exposure. 

 In continuation, four quartz test-tubes, each of about 30 c.c. capa- 

 city, were filled with part of the contents of one of the flasks, and 

 exposed at 5 cms. distance for two days. The contents were mixed 

 and used for the following experiments : 



Several attempts to obtain an osazone were made by heating with 

 excess of phenyl-hydrazine and acetic acid. The unchanged for- 

 maldehyde interfered, and although attempts were made to overcome 

 this by fractional precipitation, the most that could be obtained was 

 occasionally small microscopic tufts of crystals, mixed with oily 

 and amorphous material. 



Changes due to the exposure are evidenced by the deep orange 

 colour and orange precipitate obtained on boiling with the phenyl- 

 hydrazine in the case of the exposed solution, while similar treat- 

 ment of the unexposed formaldehyde gave only a pale yellow colour 

 in both solution and precipitate. 



