Prof. li. Hansen's Spectral- Analytical Researches. 535 



several times subjected to the same treatment with oxalic acid 

 and magnesite j the oxides are at last dissolved in sulphuric acid, 

 the acid is removed by evaporation, and the salt is heated to 

 incipient redness. In order to obtain from this salt a lantha- 

 num compound perfectly free from didymium, the original pro- 

 cess of Mosander is the simplest and safest. The dehydrated 

 sulphate is dissolved in small quantities of water at 0° to 5° C. ; 

 the liquid is warmed until the lanthanum salt precipitates in 

 the form of a soft white mass of small needles ; this mass is 

 placed in a funnel heated to 100° C, and the mother liquor 

 (which is set aside for the preparation of didymium salts) is 

 removed by means of the water air-pump. The mass dehy- 

 drated at an incipient red heat is treated in the same way six 

 or eight times. A layer 0'2 millim. thick of a concentrated 

 solution of pure lanthanum sulphate shows no trace of the 

 absorption-lines of didymium or erbium, nor, after conversion 

 into chloride and testing in the spark-spectrum, any of the lines 

 of cerium or yttrium. Samples of lanthanum oxalate, obtained 

 from this material by fractional precipitation, exhibited, when 

 converted into chlorides, no differences in the number, position, 

 and relative intensity of their lines. 



Lanthanum oxalate comports itself towards ammonium oxa- 

 late in the same manner as the corresponding cerium salt ; the 

 fractional precipitates from ammonium oxalate show the same 

 spectra throughout. Lanthanum chloride gives no flame-spec- 

 trum ; but it is characterized by a very brilliant spark-spectrum 

 rich in lines, no. 14 e } Plate V. 



The first mother liquor from the lanthanum sulphate is the 

 starting-point in the preparation of pure didymium compounds. 

 The methods hitherto put forward for the purifying of didymium 

 compounds yield substances in which lanthanum is easily 

 detected by the spark-spectrum. The removal of these impuri- 

 ties is carried out by me as follows : — If didymium sulphate, 

 prepared according to Mosander's original directions, be decom- 

 posed by oxalic acid gradually added, the oxalates of didymium 

 and lanthanum, which are at first precipitated, are again dis- 

 solved, until a point is reached when, upon the further addition 

 of acid and shaking, the formation of a permanent crystalline 

 amethyst-coloured precipitate commences. This first part of 

 the precipitate is rich in didymium ; it is separated from the 

 other part, which contains larger quantities of lanthanum. The 

 precipitate is converted into neutral dehydrated didymium sul- 

 phate, which is again treated in the way just described; and 

 this treatment is repeated until the last product does not exhibit 

 any lanthanum-lines in the spark-spectrum. 



The spark-spectrum of pure didymium chloride exhibits, as is 



