2 1 ('» I 'mi Name and Bosworth — Rates of Solution of 



Experiments with Nickel and Cobalt. 



No special tests for purity were applied to the sample of 

 nickel used, but the fact that in dissolving in acid or iodine 

 solution it retained a bright surface and gave no residue may 

 be accepted as an indication that it contained very little carbon. 

 The velocity constants obtained agree very closely with those 

 for iron and cadmium. 



The first two experiments with cobalt, Nos. 3 and 4 in the 

 table, were carried out with a comparatively impure sample 

 of the metal, which on treatment with acids or iodine solution 

 soon acquired a black, closely adherent coating. This cobalt 

 was proved by chemical tests to contain a good deal of nickel 

 and iron together with traces of copper. The results obtained 

 with this sample of cobalt are much lower than with iron or 

 nickel. « 



For experiments 5 and 6 disks of pure electrolytic cobalt 

 were used, obtained by depositing the metal on a disk of plati- 

 num of the standard diameter from a solution of carefully 

 purified cobalt chloride containing sodium formate and a 

 slight excess of formic acid. Very smooth and compact 

 deposits of ample thickness were thus obtained, using a 

 current density of 0*4 amp. per sq. dm. The purification of 

 the cobalt chloride was based on two successive precipitations 

 as potassium cobaltinitrite. 



These disks retained a perfectly clean surface in dissolving. 

 Before using they were measured with a micrometer and the 

 constants obtained were corrected for the difference in surface 

 area from that of the standard size (38-3 mm diam. and 0-5 mm 

 thick). 



Since experiments 5 and 6 were performed after the read- 

 justment of the apparatus they are to be compared with the 

 value 6 - 86 of Table II. The agreement is evidently all that 

 could be desired. 



Experiments with Tin, Magnesium, and Aluminum. 

 The results obtained with tin are of particular interest since 

 they show the nature of the disturbances to be expected in the 

 case of a metal which may give two different soluble iodides. 

 As compared with the metals previously considered the velo- 

 city constants for tin were found to be, on the average, 

 about 15 per cent higher. This can not be attributed to 

 impurities, for the sample of tin used was of the highest 

 purity which could be purchased, and was further purified 

 for this special purpose by decomposing it with nitric acid, 

 washing the metastannic acid, and finally reducing with cane 

 sugar. The probable explanation is as follows : Iodine reacts 

 with a stannous salt according to the equation 

 Sn++ + I, = Sn++++ + 2 I', 



