DIDYMIUM. 239 



didymium oxide was dissolved in nitric acid, the solution 

 was evaporated with sulphuric acid, and the weight of the 

 resulting sulphate was ascertained. I subjoin the weighings 

 and the percentage of 01263 in 01,(804)3 : 



Di^Oy Di^^SO^y Per cent. Di^Oy 



2.257 gnu. 3.844 grm. 58.715 



1.086 " 1.84S5 " 58.750 



1. 1525 " 1. 9615 " 58.756 



1.3635 " 2.319 " 58.797 



1-9655 " 3-3435 " 58-786 



1.528 " 2.599 " 58-792 



Mean, 58.766, dr .0087 



Hence Di = 146.804. If SO 3 = 80, Di = 147.021. 



This determination is undoubtedly the best of all, and 

 might properl}^ be accepted to the exclusion of the others. 

 Still, it is worth while to combine all the figures into one 

 general mean. For the percentage of Di^Og in 012(804)3 

 we have the following data : 



Marignac 58.270, ± .0115 



Erk and Hermann 58.107, ± .0112 



Zschiesche 57.926, d= .094 



Cleve 58.766, ± .0087 



General mean 58.451, ± .0059 



For the atomic weight of didymium we have now three 

 independent values : 



From per cent. Di.p., in 012(804)3 Di = 144.604, =fc .031 



" Marignac's chloride analyses " = 143.637, i .263 



" Marignac's and Erk's BaSO^ratio-- " = 143.929, dr .189 



General mean " =; 144.573, it .0306 



If = 16, Di = 144.906. 



