EVOLUTION OF GASES FROM HOMOGENEOUS LIQUIDS. 273 



the rate of evolution of gas, external conditions remaining the same, is at first slow, 

 then gradually increases until it reaches a maximum, and for some time constant, rate, 

 from which point it decreases constantly with the diminution of the mass undergoing 

 decomposition. The curve, then, representing the velocity in terms of mass would 

 thus show a point of contrary curvature. This phenomenon of initial acceleration has 

 previously been observed in reactions occurring between two gases giving a third gas 

 as a resultant, as also between two liquids with liquid resultants. Thus BUNSEN and 

 ROSCOE in their photo-chemical researches* pointed out that the combination of 

 hydrogen with chlorine, under the influence of direct sunlight, starts at first slowly, 

 but quickly increases, reaching a maximum ; this phenomenon they called " photo- 

 chemical induction." To illustrate their results, they studied the rate of replacement 

 of hydrogen in tartaric acid by bromine, both the reacting substances being in dilute 

 solution ; this rate was found to increase slowly, reaching a maximum point. Again, 

 BERTHELOT and PEAN DE SAINT GiLLEst observed the same period of initial acceleration 

 in their studies on the rate of formation of ethereal salts, less marked in the etherifica- 

 tion of acetic acid by ethyl alcohol, but most marked in the case of valeric acid. On 

 this point they write as follows^ : " Pour la concevoir il faut admettre une sorte 

 d'inertie, de resistance .-i vaincre qui retarde la combinaison dans les premiers instants, 

 et dont les effects compensent et au-del;\ pendant un certain temps ceux qui rdsultent 

 d'un e"tat de concentration plus grande dans le systeme. Cette acce'le'ration initielle 

 semble done constituer un caractere assez ge'ne'ral de ce genre de reactions." It is 

 apparently also common to the kind of reactions, here investigated, in which a gas is 

 formed from a liquid. But the intervals of time required for the unit of chemical 

 change do not, however, decrease in any regular uniformity during this period of 

 acceleration ; it is probable that the phenomenon is the result of several causes, the 

 peculiar effect of each of which cannot at present be determined by mathematical 

 analysis. In some cases it is due to a storage of gas within the liquid, particularly in 

 the formation of nitric oxide from nitric acid by means of ferrous sulphate, and in 

 others a delay is occasioned by the actual resolution of more complex into simpler 

 molecules, or even atoms, which in their turn react to form the gaseous product. 

 Thus, in the formation of nitrous oxide from ammonium nitrate, the delay may be 

 occasioned by the resolution of the molecule of the salt into ammonia and nitric acid, 

 as observed by BERTHELOT. In the case of the decomposition of potassium ferro- 

 cyanide by sulphuric acid the double salt at first splits up into the separate metallic 

 cyanides, which yield, by the action of the acid, hydrocyanic and subsequently 

 formic acid. This last is then decomposed into carbonic oxide and water. It would 

 appear that hydrocyanic acid is always an initial product in this decomposition. 

 Thus, to put the proposition in a general form, if in any chemical system a 



' Phil. Trans.,' 1857, p. 355. 



t ' Annales de Chimie,' [8], vol. 66, 1862, pp. 5-153. 

 t Ibid., pp. 26, 65. 

 MDCCCLXXXV1II. A. 2 N 



