1904] The Chemical Dynamics of Photographic Development. 451 



used, the ultimate analytical error for so small a quantity of AgCl 

 may seriously affect the results, while with larger plates the irregularity 

 in the developed density, primarily due to coating, causes errors 

 which are only partially compensated by taking many readings over 

 the plate. It would be preferable, moreover, to precipitate the silver 

 as AgBr, both owing to its slighter solubility and the greater weight 

 of precipitate obtained. 



A preliminary gravimetric determination was made, in order to 

 check the volumetric method subsequently employed. The densities 

 ranged from 2'4 to 3'25 and the value of P obtained was 0*01035. 



The volumetric method adopted was Volhard's,* with thiocyanate 

 sometimes controlled by the precipitation with KBr. The standard 

 silver solution was N/100 AgN0 3 , prepared by dissolving 1 -6997 grammes 

 pure AgN0 3 in water with excess of HN0 3 and made up to 1000 c.c. 

 at 15 C. This was checked by analysis, and two concordant deter- 

 minations gave the factor 1-001. The standard KCNS, N/100 had the 

 factor 1-017. 



Details of the Estimation on Plates. Half-plates and whole-plates were 

 exposed at 2 to 3 metres, carefully developed, fixed and washed. Very 

 careful washing was necessary in order to extract all soluble salts. 

 They were dried at the standard temperature. The density was 

 measured in 20 to 30 places, f and the area ; after this the film was 

 removed by dilute HF and soaked in many changes of distilled water, 

 with frequent pressure between filter paper, to remove all soluble salts 

 which might interfere.:]: The film was then dissolved in slight excess 

 of pure HN0 3 , SG 1*49 and the gelatine completely destroyed by 

 heating, as otherwise it interferes with the subsequent titration by 

 forming a pseudo-solution of the precipitate and retarding its aggre- 

 gation.^ The solution was made up to 25 c.c. at 15 C. and titrated 

 with N/100 KCNS, using 2 c.c. of 10-per-cent. ferric-am. sulphate as 

 indicator. 



The following series is exemplary. Four series . of measurements 

 were made in this way, the density ranging from 0*50 to 3'5, and the 

 curve shows that there is no bias throughout the range. 



Series IV. 



* Cf. Mohr's ' Titrirmethode.' 



t This was necessary, as variations in the developed density up to 10 per cent, 

 occurred, chiefly due to coating errors. 



J Cf. Lumiere and Seyewetz " On Eetention of Hypo," Eef. in P. J.,' 1902. 

 Lobry de Bruyn <Ber.,' 1902, vol. 35, pp. 3079 to 3082. 



