246 PROCEEDINGS OP THE AMERICAN ACADEMY. 



ferrous bromide contain no chlorine, the salt obtained in the above 

 manner gives a barely perceptible coloration with the sulphocyanate, 

 showing that reduction is complete. On the other hand, if the ferrous 

 bromide was prepared from commercial bromine, it was never found 

 possible to eliminate completely the ferric salt; although by colorimetric 

 comparison with standard ferric solutions to which amnionic sulpho- 

 cyanate had been added, it was shown that the proportion of iron in the 

 ferric state did not amount on an average to more than two hundredths 

 of one per cent. The greater difficulty with which ferric chloride is 

 reduced is doubtless the cause of this difference in behavior of the two 

 varieties of material. 



The apparatus with which these operations were carried out was 

 essentially that used for similar purposes in the determinations of the 

 atomic weights of cobalt,* nickel, f and uranium $ in this laboratory. A 

 mixture of air and ammonia was passed over heated rolls of copper 

 gauze and the excess of ammonia was removed by means of sulphuric 

 acid. The partially dried gas, after bubbling through bromine in a 

 small flask, was conducted through hydrobromic acid in which was sus- 

 pended red phosphorus to convert the bromine into hydrobromic acid. 

 Suitable desiccating agents then dried the mixture of gases. Pure iron, 

 contained iu an unglazed porcelain boat, was placed in a porcelain tube 

 heated by a Fletcher furnace. The mixed nitrogen and hydrobromic 

 acid gases were passed over the hot iron, forming ferrous bromide, and 

 since the temperature was high enough to sublime the ferrous bromide, 

 the salt was carried along by the current of gases to be deposited in a 

 smaller cool porcelain tube which telescoped into the larger tube. As 

 soon as the inner end of the smaller tube was stopped up with bromide, 

 the furnace was allowed to cool for some time, and the sublimed salt 

 while still warm was transferred to a weighing bottle. The portions 

 next the porcelain were never collected for analysis. The ferrous bro- 

 mide, contained in a platinum boat, was then heated to about 400° C. in 

 a current of nitrogen and hydrobromic acid gases for an hour or more. 

 Finally, after the boat had become cool, these gases were displaced, 

 first by nitrogen, then by dry air, and the boat was transferred to a 

 weighing bottle without exposure to moisture, by means of the bottling 

 apparatus so frequently used for the purpose in this laboratory. § 



* Richards and Baxter, These Proceedings, 33, 117. 

 t Richards and Cushman, These Proceedings, 33, 99. 

 J Richards and Merigold, These Proceedings, 37, 378. 

 § Richards and Parker, These Proceedings, 32, 59. 



