CHEMISTRY, 



343 



It was definitely proved by direct analysis that the insoluble sub- 

 stance formed when praseodymium chloride is fused in a current of 

 hydrochloric-acid gas is a basic chloride instead of an insoluble modifi- 

 cation of the chloride, as was supposed at first. By careful drying of 

 the chloride previous to fusion, and by efficient drying of the hydro- 

 chloric-acid gas in which the salt was fused, the quantity of insoluble 

 substance was always kept as low as a few tenths of a milligram, and 

 in some cases none whatever could be seen. In any case the error 

 involved is very small. 



The analyses were carried out in the usual way by comparison of the 

 anhydrous salt with pure silver and by weighing the silver chloride. 

 The results of the analyses are shown in the tables. Fraction 3474 

 was the purest fraction in one of the later series of crystallizations of 

 the double ammonium nitrate. Fraction 4383 was the most. Fraction 

 4368 the least soluble of the last series. 



The close similarity of the different fractions is very satisfactory in 

 showing that, so far as fractional crystallization of the double ammo- 

 nium nitrate is concerned, no further separation or purification of the 

 praseodymium is to be expected. The atomic weight of praseodymium 

 seems to be very close to 140.92, a value 0.3 unit higher than the one 

 adopted by the International Committee on Atomic Weights, but in 

 fair agreement with some of the most recent work upon this constant. 



The Atomic Weight of Neodymium. 



Mr. Stewart also made additional analyses of the fractions of neo- 

 dymium chloride previously analyzed by Mr. Whitcomb (see Year 

 Book No. 11). As in the case of praseodymium chloride, the neo- 

 dymium chloride, even if very carefully dehydrated by efflorescence at 

 low temperatures, when fused in dry hydrochloric-acid gas frequently 

 yields a very small proportion of insoluble material, which Mr. Stewart 

 proved by direct analyses to be the basic chloride. In the analyses 

 of neodymium chloride made by Mr. Stewart this difficulty was 

 avoided. The following table compares the results of the two experi- 

 ments : 



As has already been stated (Year Book No. 11), Fraction 1+2+3 

 contains a trace of samarium, which would raise the apparent atomic 

 weight of neodymium, while Fractions 12 to 15 contain praseodymium, 



