A REVISION OF THE ATOMIC WEIGHTS OF IODINE AND SILVER. I35 



DISCUSSION OF RESULTS. 



The average of this series is higher than that of Series I by less than one one- 

 thousandth of a per cent, the atomic weight of silver calculated from the ratio 

 of silver to pentoxide being 107.850. While it is true that impurity in the iodic 

 acid not containing halogens would tend to lower the observed atomic weight 

 of silver, the close agreement of the two series carried out with material of 

 fairly diverse nature practically eliminates impurity in the iodic acid or silver 

 as the cause of the low resulting value for the atomic weight of silver. Sample 

 I of iodic acid was crystallized only three times from aqueous solution, while 

 Samples II and III were both crystallized at least ten times. It is improbable 

 that any impurity could have passed through the additional crystallizations 

 without appreciable diminution in quantity. It has already been shown that 

 mineral impurities were surely absent even before the crystallization of the 

 acid. Nitric acid could hardly have survived the prolonged heating at 240° 

 even if it had not been completely removed by the many crystallizations of the 

 iodic acid. 



With regard to impurities containing halogens other than iodine, it may be 

 pointed out that Samples I and III were prepared from iodine and nitric acid 

 which had been very thoroughly freed from chlorine and bromine, while even in 

 the case of Sample II only traces of chlorine could have been present. Aside 

 from these facts it is decidedly improbable that an oxygen acid of either chlorine 

 or bromine could have been formed and then have accompanied the iodic acid 

 during its purification, for during the heating of the acid both before and after 

 crystallization such impurities would have been either volatilized or destroyed. 

 Impurity of chlorine would probably tend to lower the observed atomic weight 

 of silver on account of the relatively high solubility of silver chloride. For 

 the same reason a trace of chlorine in either the hydrazine or the nitric acid em- 

 ployed in the analysis would have had no injurious effect. Impurity of bromine 

 would produce the reverse effect if present in the form of a compound analogous 

 to iodine pentoxide. 



The presence of a halogen of higher atomic weight than iodine, forming an 

 insoluble silver salt, if present as pentoxide would lower the observed atomic 

 weight of silver. The existence of such an element is purely hypothetical, 

 however, and what evidence exists is contrary to such an hypothesis. One of 

 us has recently searched for such an element in vain.^ 



Two other possible contingencies must be considered, the presence of either 

 free iodine or oxides of iodine higher than the pentoxide. Free iodine might 

 result from reduction of the pentoxide during the heating. Such reduction or 

 decomposition actually does take place to an extremely slight extent when the 



1 Baxter: Proc. Anier. Acad., 40, 422 (1904); Jour. Amer. Chem. Soc, 26, 1580; Zeit. 

 anorg. Chem., 43, 17. (See page 93.) 







