ON ORGANIC COMPOUNDS 89 



two tubes were placed in equally favourable positions alongside each 

 other at about 5 cms. distance from the quartz tube of a mercury- 

 vapour lamp. At the outset the formaldehyde gave no reduction of 

 Benedict's solution. After three hours' exposure the contents of the 

 quartz tube gave a fair reduction of Benedict's solution, while those 

 of the glass tube still gave a complete negative with this test solu- 

 tion. At the expiry of six and a half hours' exposure the contents of 

 the quartz tube gave a copious reduction, but the contents of the 

 glass tube had not yet developed any reducing substance. At the 

 end of twelve hours the contents of both tubes showed reduction, 

 those of the glass tube to about the same extent as had been obtained 

 in the quartz tube in three hours, while the contents of the quartz 

 tube now gave a heavy precipitate of cuprous oxide when tested by 

 Benedict's solution. 



The remainder of the contents of the quartz tube were distilled 

 at ordinary atmospheric pressure almost to dryness, leaving a white 

 solid residue in the distilling flask. The residue was taken up with 

 warm water, in which it readily dissolved, and made up to the 

 original volume. Both distillate and residue were then tested 

 with Benedict's solution, and both were found to reduce it readily, 

 the residue containing apparently somewhat more reducing sub- 

 stancs than the distillate. Again, both distillate and residue 

 reduce neutral silver nitrate solution and mercuric chloride solution. 

 This experiment shows that the short wave-lengths are the most 

 powerful, and that the limit lies at about the shortest wave-lengths 

 transmissible by glass. 



Experiment VII. In a somewhat similar experiment two quartz 

 test-tubes were taken, of the same dimensions as above, and the 

 same concentration of formaldehyde was employed, but around one 

 of the tubes as thin a sheet of mica as could be split off was folded 

 so as completely to surround the quartz, and held in position by two 

 loops of thread at top and bottom. The mica was so thin that it 

 readily folded over the test-tube (about 3 cms. in diameter) without 

 cracking. Yet this thin layer of mica was so effective a screen that 

 there was not a trace of reduction after a twelve hours' exposure 

 similar to that given above in glass ; the contents of the unshielded 

 quartz tube showed a distinct reduction after two hours' exposure. 



Experiment VIII. Only preliminary experiments have hitherto 

 been made towards the identification of the substance or substances 

 giving the reductions with these metallic salts. It is natural to 



