POTASSIUM, RUBIDIUM, AND C/ESIUM. 15 



DESCRIPTION OF MATERIALS. 

 LITHIUM. 



Three different bars of lithium obtained at different times were used for 

 the determination of the compressibility of this metal. That used in the 

 preliminary samples was slightly impure, containing either sodium or 

 potassium, as was indicated by its specific gravity and somewhat too large 

 compressibility. On the other hand, that used in the final determinations 

 was very pure, giving neither sodium nor potassium lines in the spectro- 

 scope, and containing only 0.3 per cent of aluminum and 0.1 per cent of 

 iron, according to a careful analysis made by F. N. Brink, of this labo- 

 ratory, who analyzed the sodium and potassium also. The density was 

 found by him to be only 0.534, as will be shown in a further communica- 

 tion. The impurity present was too small in amount to exercise a signifi- 

 cant effect on the compressibility. 



SODIUM. 



The sodium employed was the commercial article prepared by the Mal- 

 linckrodt Chemical Works of St. Louis. It was found to contain 0.035 

 per cent of iron, and 0.03 per cent of magnesium ; but no other impurities 

 in significant amounts were detected. This small proportion of contam- 

 ination could not have affected its compressibility. The inconsiderable 

 weight of hydrogen, which almost always exists in commercial sodium, was 

 found by Brink to have no essential effect on the density of the metals, 

 and therefore this also could hardly have affected the compressibility. 

 The density was found to be 0.972, in confirmation of Brink's results upon 

 the same preparation of sodium. 



POTASSIUM. 



The potassium was a German preparation of a very high degree of 

 purity, containing only minute traces of iron and aluminum, and 0.03 per 

 cent of magnesium. No other impurities were found, except hydrogen, 

 as in the case of sodium. The metal was fused under a limpid paraffin 

 oil, and was subsequently cut and shaped to fit as closely as possible into 

 the jacket. Its density was 0.862. 



RUBIDIUM. 



This metal the commercial article furnished by the same German firm 

 which supplied the potassium was very impure, in marked contrast to 

 the potassium. It had been our intention to return to rubidium, preparing a 

 pure specimen ourselves, after the work on caesium, but because the latter 

 work seemed to leave no doubt that the compressibility calculated from the 

 results with the impure metal were approximately correct, and time was 



