BAXTER AND CHAPIN. — ATOMIC WEIGHT OF NEODYMIUM. 219 



is weakened considerably by lanthanum and cerium ammonium nitrates 

 and other salts, as well as by concentrated nitric acid, the effect of 

 which was mentioned in Auer's earlier paper. Clearly any quantitative 

 comparison of the intensity of the praseodymium spectrum from solu- 

 tions of praseodymium salts alone, with that from solutions consisting 

 mainly of other rare earth salts, is unreliable. 



Feit and Przibylla ^ dissolved weighed amounts of various rare earth 

 oxides in sulphuric acid and titrated the excess of acid, using methyl 

 orange as the indicator. The endpoint was interfered with by the pink 

 color of the neodymium salt, but this difficulty was obviated by com- 

 paring the solutions with standards containing the same amounts of 

 indicator and neutral neodymium salt. Their neodymium material 

 was prepared by fractional crystallization of the double magnesium ni- 

 trate. This preparation showed no trace of the strongest praseodym- 

 ium and samarium absorption bands, although the " hellblau " color 

 of the oxide was considered the best proof of its purity. Their result 

 for the atomic weight of neodymium was 144.60. 



Holmberg2 after a study of the rare earth salts of twelve organic 

 acids, selected the metanitrobenzol sulphonate as particularly suitable 

 for the purification of neodymium, and by one hundred and sixty-two 

 series of crystallizations obtained thirteen fractions ranging in atomic 

 weight from 144.0 to 145.3. The purest fractions were selected and 

 the oxide obtained by ignition of the oxalate was weighed and con- 

 verted into sulphate. The value 144.10 was obtained as the mean of 

 six results varying from 144.03 to 144.15. 



The International Committee on Atomic Weights have chosen as the 

 most probable value 144.3, which represents an average of the more 

 recent determinations. 



The Separation of Neodymia from Other Rare Earths. 



It is a well-known fact that when a salt of a mixture of rare earths 

 is subjected to fractional crystallization, the less soluble salts tend to 

 concentrate at the less soluble end of each series of fractions, and the 

 more soluble salts at the opposite end of the series. In most cases it 

 is less difficult to free an earth very thoroughly from more soluble im- 

 purities than it is to eliminate impurities which tend to concentrate in 

 the crystals. An obvious although not generally recognized way out 

 of the difficulty is to crystallize the substance in such different forms 

 that impurities which tend to concentrate in the crystals in one case, 



1 Zeit. anorg. Chem., 43, 202 (1905). 

 ' 2 ibi(j^ 53^ 83 (1907). 



