RICHARDS. — ATOMIC WEIGHT OP STRONTIUM. 381 



weighed as the baric bromide had been. The decomposition of the 

 salt was so great, however, that the uncertainty of the alkalimetric 

 correction sometimes amounted to two or three tenths of a milligram ; 

 lience this method was clearly inadmissible. 



The fusion of the salt in a platinum boat in a stream of nitrogen 

 gave much better results, and two or three further preliminary deter- 

 minations by this method gave promise of much greater accuracy. 

 It is probable that the slight decomposition which occurred even in 

 the atmosphere of nitrogen was due to the presence at 250°-300° of a 

 slight trace of moisture. 



The presence of an excess of hydrobromic acid must necessarily 

 lessen or prevent this decomposition ; hence in three succeeding deter- 

 minations (Nos. 13, 14, 15, below) pure dry hydrogen bromide was 

 added to the nitrogen in which the combustion was conducted. 



In these cases, however, the platinum boat, which had previously 

 remained quite constant in weight, was evidently attacked, since upon 

 one occasion (Exp. 15) it lost over two tenths of a milligram, and the 

 pure white strontic bromide became tinged with a brown color. The 

 weight of the boat after each fusion was taken as the true weight, 

 because the bromide of platinum, if formed, must precipitate nearly as 

 much silver as the bromide of strontium. 



In order to avoid the corrosion of the boat, hydrogen was added in 

 small quantities to the mixture of gases. This, by preventing the 

 dissociation of the hydrobromic acid, effectually preserved the platinum, 

 and the boat remained constant in "weight. The pure translucent or 

 transparent colorlessness of the fused salt left nothing to be desired. 

 A somewhat complex piece of apparatus was needed for the purpose. 

 (See page 382.) A mixture of six volumes of pure nitrogen (made 

 by passing air and ammonia over red-hot copper) and one volume of 

 pure hydrogen was delivered from a gas holder through a succession 

 of tubes of red-hot copper, dilute chromic and sulphuric acids, concen- 

 trated alkaline pyrogallol, and fused potash, into the arrangement for 

 preparing hydrobromic acid. This, as well as all the apparatus fol- 

 lowing, was without rubber connections, the ground joints being made 

 tight by means of syrupy phosphoric acid (Morley) and flexible by 

 means of fine glass gridirons (Finkener). The pure dry nitrogen and 

 hydrogen were led in the first place into a flask containing bromine, 

 and then over asbestos and red phosphorus saturated with pure 

 fuming hydrobromic acid. The bromine and hydrobromic acid were 

 proved to be pure by the usual quantitative analysis, and the red 

 phosphorus was ground and washed many times with pure water to 



