BAXTER AND STEWART. — PRASEODYMIUM CHLORIDE. 195 



added to that obtained by the spectroscope, that all the fractions 

 analyzed were of a high degree of purity. 



Fraction 4383, which contained all the neodymium which may have 

 accumulated in forty-one series of crystallizations, gives a slightly 

 lower instead of higher figure than the average, while fraction 4368 

 which contained more cerium than any other fraction gives a result 

 identical with the average. This close similarity is very striking when 

 one considers that a range of sixteen fractions is included in those 

 examined. Fraction 3474, selected at an earlier stage of the fractiona- 

 tion as being very pure, also shows no sign of being different from the 

 other fractions. 



The foregoing results leave little opportunity for choice. There 

 seems to be no reason for preferring the value obtained with any one 

 or more fractions to those resulting with the others. Possibly the 

 average of Analyses 5, 6, 9, 10, 16, 24, 25, 30, 31, 38 and 39, 140.913, 

 is more reliable than the average of all, for in these analyses the 

 praseodymium chloride dissolved rapidly and completely at the start, 

 and was probably entirely free from basic salt, but we believe that the 

 final corrected average of all the experiments, 140.92, represents fairly 

 the best material we succeeded in preparing. This result is not far 

 from the value obtained by Brauner in both his researches, but is over 

 three tenths of a unit higher than the choice of the International 

 Committee Upon Atomic Weights. 



We are indebted particularly to the Carnegie Institution of Wash- 

 ington for pecuniary assistance in carrying out this investigation, 

 and also to the Cyrus M. Warren Fund for Research in Harvard 

 University for indispensable platinum vessels, as well as to Dr. 

 H. S. Miner of the Welsbach Light Company for the praseodymium 

 material. 



Cambridge, Mass. 



