KLISHA MITCHELIv SCIENTIFIC SOCIETY. 63 



dissolved in the smallest possible quantity of dilute 

 li3Tlrochloric acid and sodium iodate (NaI03) added 

 in excess. The solution is heated for about fif- 

 teen minutes. It is then allowed to stand twelve hours 

 filtered, washed down with boiling- water, dissolved in 

 hvdrochloric acid and finallv precipitated with ammon- 

 ia, ignited and weig^hed." I found in an analysis 

 0.0515 g-.ZrO, when I had used 0.0520 g. 



Analyses were made also according to his recom- 

 mendation of the use of from five to ten per cent, of 

 sodium chloride. The results obtained were high, 

 doubtless due to imperfect washing. An example : 



Found. Used. 



ZrO, 0.0595 0.0520 



The numerous experiments made served mereh' to 

 confirm Davis' work. It was necessar}^ to avoid a too 

 far neutralization with sodium carbonate as the 

 separated zirconium was contaminated with varying 

 amounts of aluminium. The permanent precipitate 

 formed b}^ the sodium carbonate was diificult at times to 

 redissolve in a small amount of dilute hydrochloric acid. 

 Yet an excess of acid must be avoided, for it was learned 

 by experiments, as Davis had noted, that the presence 

 of even 0.1 per cent, bv ^veight of hydrochloric acid 

 would cause low results. Four hours was a sufficient 

 time for complete separation however. An experiment 

 with the sulphate solution showed no action whatever. 

 Kven the small amount of sulphuric acid in an alumin- 

 ium sulphate solution was found to interfere, hence the 

 necessity of having- a hydrochloric acid solution, free 

 from sulphuric acid, was apparent. Davis evidently 

 noted this as he was particular in having' a pure solu- 

 tion of aluminium chloride in his experiments. 



