422 PROCEEDINGS OP THE AMERICAN ACADEMY. 



sufficient to free this element from chlorine,* three distillations of iodine 

 from a solution of an iodide should insure complete elimination of chlorine 

 and bromine even if the original substance were very impure. The puri- 

 fied iodine was again converted into hydriodic acid by hydrogen sulphide, 

 and this into ammonium iodide by an excess of freshly distilled ammonia. 

 This material was used in the analyses made with Sample I. 



It was suggested by Professor T. W. Richards that an undiscovered 

 element resembling the halogens, but of a higher atomic weight, might 

 exist, and an attempt to detect the existence of such an element was 

 carried out as follows : One pound of iodine was purified as described 

 above, but the process of converting the iodjne into hydriodic acid by 

 hydrogen sulphide, and the hydriodic acid into iodine by potassic 

 permanganate, was repeated four times. The iodine obtained in the last 

 distillation with permanganate, amounting in all to about fifty grams, 

 was set free in four fractions by introducing the permanganate in portions 

 of about two grams each. An element belonging to the halogen family, 

 and of higher atomic weight than iodine, should be more easily set free 

 than iodine, hence the first of the four fractions should have contained 

 nearly all of such an element which occurred in the pound of iodine. 

 Each one of the fractions was once distilled with pure water, and was 

 then converted into ammonium iodide, as in the case of Sample I. The 

 first fraction was Fraction 1 of Sample II, the fourth was Fraction 4 of 

 Sample II. No difference in appearance between these samples and 

 Sample I or even the original crude iodine could be detected. 



' All reagents used in the course of the purification or in the analyses 

 were carefully purified. Acids and ammonia were redistilled, usually 

 with the use of a platinum condenser. The water was distilled twice, 

 once from alkaline permanganate, once from a solution containing a 

 trace of sulphuric acid. In both distillations block tin condensers were 

 used. 



Method of Analysis. — In order to convert a known weight of silver 

 into silver iodide the following method was employed. The silver was 

 dissolved in nitric acid in an Erlenmeyer flask, into the neck of which a 

 column of glass bulbs was ground. t The acid consisted of the constant 

 boiling distillate diluted with an equal volume of water. Solution was 

 effected by warming so gently that almost no gas was evolved. When 



the silver was completely dissolved, the solution was diluted with an 



* Proc. Araer. Acad., 28, 17 ; 31, 165; 33, 10(5; 39, 249. 

 t Richards, Proc. Anier. Acad., 25, 198. 



