214- Van .X<n,t< and Bosworth — Rates of Solution of 



atmosphere of carbon dioxide, obtained by conducting a rapid 

 current of the gas into the upper part of the reaction vessel 

 throughout the experiment. Blank tests showed that this 

 method, when properly carried out, reduced the oxidation to a 

 negligible amount. In one case, however, experiment f>, a 

 failure of the generating apparatus allowed some oxidation to 

 occur, the amount of which was roughly estimated, and a cal- 

 culated correction of - 07 added to the final result. The value 

 of " K (corr.)" for experiment 6 as given in the table includes 

 this correction. 



Experiments with Iron. 



The disturbing effect of impurities in the metal used is well 

 illustrated by the experiments with iron recorded in Table III. 

 In experiments 1 and 2 the disks were of common sheet iron, 

 which in dissolving set free a considerable quantity of carbon. 

 Swedish iron, supposed to be of fairly high purity, was used in 

 experiments 3 and 4. The rest were carried out with " Amer- 

 ican Ingot Iron,"* which, according to the advertised analysis, 

 is about 99'9 per cent pure. That this last material was com- 

 paratively free from carbon was evident from the bright 

 surface which it maintained in dissolving, and the clear solu- 

 tion produced. 



It is seen in the table that the common and Swedish iron 

 gave higher constants than the purer metal. A possible effect 

 of the presence of carbon in the iron is illustrated by the 

 results of certain experiments in which metallic contact existed 

 between the disk and its platinum holder. This occasonally 

 happened owing to the wearing through of the layer of 

 pyroxlin on one of the arms supporting the disk, and was 

 indicated by an abnormally high rate of solution as shown by 

 the two following experiments with cadmium in the presence 

 of barium iodide : 



A. 



9'22 



9-13 



9-20 



9-04 



9-05 



9-25 



Av. 9-15 



B. 



8-73 



8-49 



8-82 



8-21 



8-27 





Av. 8-50 



The normal value of the constant under the prevailing condi- 

 tions was 6*45. In both cases a plainly visible point of contact 

 between the cadmium and platinum was found. 



It is easy to account for the high constants in such cases. 

 Contact between the two metals establishes a short-circuited 

 cell of which platinum is the cathode. This affects the results 

 in two ways : first, the nascent hydrogen evolved on the plati- 

 num reduces iodine, thus causing an apparent increase in the 

 reaction velocity ; second, the rate of movement of the iodine 

 diffusing toward the disk in the form of 1/ ions is accelerated 

 by the superposition of an electrical force upon the osmotic 



* From the American Rolling Mill Co., Middletown, Ohio. 



