Photochemical Formation of Formaldehyde in Green Plants. 227 



The reaction employed for the purpose of estimation was, in its first form, 

 originally described by Eimini, who showed that when formaldehyde solu- 

 tions were treated with phenylhydrazine hydrochloride, and a drop of ferric 

 chloride and concentrated sulphuric acid were subse(iuently added, a 

 brilliant fuchsine-like colour developed. The reaction in this form was found 

 to be uncertain ; in the presence of too small a quantity of ferric chloride, 

 the full colour due to the reaction was not developed, whereas in the 

 presence of an excess of the reagent, the colour was rapidly destroyed. The 

 use of concentrated sulphuric acid was also inconvenient when the reaction 

 was employed quantitatively. 



The reaction is due to the formation of a condensation product of formal- 

 dehyde and phenylhydrazine, which, on oxidation, yields a weak base ; the 

 latter, in the presence of excess strong acids, yields salts, which readily 

 undergo hydrolytic dissociation on dilution. By substituting for ferric 

 chloride an oxidising agent, which, when added in excess, does not destroy 

 the colour, and by employing concentrated hydrochloric acid instead of 

 sulphuric acid for the formation of the coloured salt, the Rimini reaction 

 can be adapted to quantitative estimation of formaldehyde. In this form one 

 part of formaldehyde in one million can be readily detected, whereas Eimini 

 claimed that his original test was only sufficiently sensitive to detect one 

 part in 50,000. 



The test, with the above modifications, is carried out in the following way. 

 To 10 c.c. of the formaldehyde-containing liquid are added 2 c.c. of a 

 1-per-cent. solution (freshly made up and filtered) of phenylhydrazine 

 hydrochloride ; 1 c.c. of a fresh 5-per-cent. solution of potassium ferri- 

 cyanide is then added. On addition of 5 c.c. concentrated hydrochloric 

 acid, a brilliant colour is developed in presence of formaldehyde ; by com- 

 paring the depth of colour with that of standard solutions (of concentrations 

 between 1 in 1,000,000 and 1 in 100,000), the quantity of formaldehyde 

 present can be determined. 



By the test in the above form, free formaldehyde can be detected and, 

 when sufficiently dilute, quantitatively estimated. By a slight modification 

 formaldehyde in a polymerised form, as in the so-called oxymethylene, or 

 in the form of stable combination, as in hexamethylenetetramine 

 (urotropine), can also be detected and estimated. If to a weak solution 

 (10 c.c.) of hexamethylenetetramine be added phenylhydrazine hydro- 

 and ferri-cyanide solutions, and then hydrochloric acid in the quantities 

 given above, only a faint coloration (if any) will develop. If, however, 

 the mixture be warmed for a short time, after addition of the phenyl- 

 hydrazine hydrochloride solution, or be allowed to stand, and the ferri- 



