170 



It is not easy at first to- see why under any conditions this reac- 

 tion sliould lead to a prodnction of .sugar, nor woidd it in all 

 probability if it were a simple matter of a reaction between for- 

 maldehyde and an alkali. Bearing in mind the fact that when a 

 previously measured quantity of dilute alkali is run into lx)iling 

 dilute fornialdeliyde, the reaction is completed, and sugar is jjro- 

 duced Avithin two or three minutes, or even more rapidly if an 

 excess* of soda is used, whereas tlie same solutions kept at 12 to 15°C. 

 for two months or more develop little or no sugar, it seems probable 

 that this differenxie can liardly be due wholly to the influence of tem- 

 perature on the rate of chemical reaction. 



If a strong solution of formaldehyde is Iwiled down to 1 or \ 

 its bulk, and cooled, it solidifies to a wliite waxy mass of the poly- 

 hydrate of formaldehyde. None appears Avhile the liquid is boil- 

 ing, because of its low melting point. When a litre of 1 or of 2% 

 formaldehyde is boiled nearly to dryness, and then cooled, it leaves 

 a considerable solid residue of the polyhydrate of formaldehyde. 

 Hence, in spite of the loss of formaldehyde vapour, a dilute solution 

 can be concentrated by boiling. If the solution is evaporated at 

 a lower temperature, or under reduced pressure all, or nearly all 

 the formaldehyde escapes, and no residue is left. Presumably 

 therefore, in boiling water, the substance exists mainly as the poly- 

 drate of formaldehyde, or as paraformaldehyde, which, on cooling, 

 partially dissolves and partially dissociates to formaldehyde, so 

 long as it has not separated out in mass. Hot water poured into 

 paraformaldehyde or the solid polyhydrate soon acquires a smell of 



