266 PROCEEDINGS OP THE AMERICAN ACADEMY 



effected in each case. The process appears to deserve attention, and 

 to compare very favorably with separation by means of basic nitrates. 

 The details of this method are as follows. 



The mixed oxides in as pure a state as possible are to be dissolved 

 in a small excess of pure chlorhydric acid; the solution is then to be 

 evaporated at first on a water-bath, and then on a sand-bath, until a 

 thick syrup is obtained. This is to be transferred to a porcelain cru- 

 cible, which is then to be heated gently in a muffle. The heat should 

 not be allowed to exceed low redness. Much chlorhydric acid is given 

 off during the heating, and a white mass remains which consists in 

 part of a mixture of oxychlorides, and in part of unaltered chlorides. 

 The mass is then to be treated with hot water in rather small portions 

 at a time, the solution being poured off before each new addition. 

 When the washing is nearly complete, the solution becomes turbid. 

 The white insoluble mass and supernatant liquid are then to be evap- 

 orated to dryness, or to a thick syrup, after the addition of enough 

 chlorhydric acid to effect complete solution. The liquid poured off and 

 containing the soluble chlorides is also to be evaporated to dryness, 

 but without addition of chlorhydric acid. We have then two separate 

 portions, of which one, A, contains the soluble or relatively acid 

 chlorides, and the other, B, the neutralized oxychlorides. Each of 

 these portions is to be treated in the muffle as in the first case. In 

 this manner two new basic chlorides, B 2 and B 3 , and two new acid chlo- 

 rides, A 2 and A 3 , are obtained. The same processes are to be repeated 

 as long as the material lasts. The atomic mass of the oxides taken 

 must first be determined, and afterward the atomic masses of the 

 portions A 1} A 2 , A 3 , etc., and B 1} B 2 , B 3 , etc. This enables us to 

 determine the rate of change in the atomic masses produced by the 

 successive operations. Of course it is only necessary to take a small 

 portion of each product A l5 Bj, etc., for the determination of the 

 atomic mass. The whole process will perhaps be rendered more 

 clear by means of a diagram, and in illustration I shall select the re- 

 sults of actual work in a particular case. The starting point in this 

 case was a mixture of oxides from Texas gadolinite having the atomic 

 mass 107.5. The data are as follows : — 



(41) 0.3577 gr. gave 0.1357 gr. C 2 3 = 37.91 per cent. 



(42) 0.324G gr. " 0.1228 gr. " = 37. SI 



(43) 0.3999 gr. " 0.1515 gr. " = 37.86 " 



(44) 0.0113 gr. " 0.2814 gr. R 2 3 = 46.04 



(45) O.G670 gr. " 0.3064 gr. " = 45.94 " 



(46) 0.6303 gr. " 0.2892 gr. " = 45.88 " 



Atomic mass 107.5. 



