462 Messrs. S. E. Sheppard and C. E. K. Mees. [Dec. 20, 



ment retarded,* but the writers found that the addition of small 

 quantities of ferric oxalate, the other reaction-product, had no effect 

 comparable with that due to equimolecular quantities of bromide. 

 Only with very large concentrations of ferric oxalate (above N/25) 

 was an effect perceived. Bromide is therefore shown by this method 

 of " isolation " to have an effect per se. A more probable theory is that 

 bromide acts by lowering the concentration of the silver ions in accord- 

 ance with the reversible ionic equilibrium 



A + g + Br zz AgBr ^ AgBr, 

 (solut.) (solid) 



when we have CA S Cur = constant. 

 Now, the chemical equation gives 



dD/dt= K.C Ag .C 



for the reaction velocity, and in consequence, the velocity should be 

 inversely as the concentration of the bromide ions. That it has a 

 smaller effect than this confirms the view that the reaction-velocity is 

 chiefly determined by the diffusion equation. Nevertheless, bromide 

 may so slow the chemical reaction that its velocity begins to affect that 

 of the total reaction. 



We may further notice, as was remarked above, that the expression 



* log * = K does not hold over so wide a range when bromide is 



^ l^oo ~ "^ 



introduced, but that the values of K rise at first, and then reach a 

 fairly constant period, the induction-period apparently depending both 

 on the concentration of the bromide and on the value of D^. The 

 phenomena observed are similar to those in "graded " reactions taking 

 place in homogeneous systems.! These initial disturbances are inti- 

 mately connected with the disturbance of the density ratios by bromide. 

 In order to investigate the action more fully, a series of exposures on 

 the same plate, thus giving different values of D^, were obtained by 

 means of the previously described sensitometer ; the results of this 

 work will be noted later, with an elaborated theory of the mechanism 

 of development. 



Age of Plate. In the course of this work, it was found that the 

 velocity of development depends largely on the age of the plate, 

 diminishing with keeping. This change is accelerated by heat and 

 atmospheric conditions. This was discovered when plates of Batch I 

 were used a considerable time after making, and gave quite divergent 

 velocities according as they had been kept. The phenomenon, while the 

 cause was still unsuspected, caused much trouble. The following tables 

 exhibit the divergencies obtained : 



* Hurter and Driffield, ' Phot. Journal, 1 1898; Luther, ' Chemische Yorgange in 

 der Photographic,' 1900. 



f J. Walker, ' Proc. Eoy. Soc., Edinburgh,' December, 1897. 



