[ 119 ] 



XY. On the Theory of Fractional Precipitation. 

 By John J. Hood, D.Sc* 



THERE are at the present time several elements known 

 belonging to the group of earths, analogous in some 

 respects both to alumina and to lime, resembling each other 

 so closely in their chemical properties that their separation 

 seems to be one of the most tedious operations in mineral 

 chemistry. So similar, indeed, are these elements in their 

 chemical behaviour, that many, if not all, of the recent additions 

 that have been made to the list of earths have passed through 

 the hands of several eminent chemists without raising a sus- 

 picion that the material whose chemical properties they were 

 examining was in reality a complex body. For instance, 

 what was up till about the year 1877 termed erbia, an oxide 

 found in gadolinite and a few other minerals, is now known 

 to be a mixture of no less than five oxides — viz. scandia, 

 ytterbia, holmia, thulia, and the true erbia. 



The investigations that resulted in these discoveries may be 

 considered as extensions of the work of Mosander, who in 

 1843 showed that the gadolinite earth, then termed yttria, 

 was a mixture of three oxides — yttria and other two which he 

 named terbia and erbia, separable from each other by frac- 

 tional precipitation as oxalates. 



A similar history belongs to the earths extracted from the 

 mineral cerite ; the latest discovery in this direction! being 

 that the didymia from this mineral contains a considerable 

 proportion of samaria, an earth first discovered in the mineral 

 samarskite in 1878 by Delafontaine. 



The methods that are employed for the separation of these 

 earths from each other depend for success upon differences 

 existing in the basic strengths of the several oxides ; that is 

 to say, if a mixture of salts, particularly the nitrates of two 

 or more of these oxides be heated, there will be a tendency 

 for the nitrates of some of the oxides, the less basic, to decom- 

 pose first or in relatively greater quantity than the nitrates of 

 the more basic oxides : or, if to a dilute solution of these salts 

 a precipitant be added such as ammonic hydrate, less in 

 amount than is requisite to precipitate all the salts, some of the 

 hydrates or oxides will tend to accumulate more rapidly in the 

 precipitates than others by many repetitions of the process. 

 If differences do exist in the basic strengths of the materials, 



* Communicated by the Author. 



t According to Welsbach, didymium is a mixture of two elements 

 which he names neodymium and praseodymium (Journ. Ohem. Soc. Nov. 



1885). 



