290 Weidman and Zenker — Marignaeite. 



only closely approximate, for as shown by Hall,* Smith,t and 

 Warren,;}: it is impossible by present methods to obtain a perfect 

 separation of these two elements. Six crystallizations were 

 made to separate the columbium and tantalum. 



The earths after being precipitated together were treated 

 in acid solution with ammonium oxalate. The oxalates were 

 ignited, dissolved in nitric acid, and the excess of acid removed 

 by evaporation. The nitrates were treated with a saturated 

 solution of potassium sulphate. The thoria was separated by 

 means of the thiosulphate and ammonium oxalate methods. 

 The ceria was purified by suspending the hydrate in a solution 

 of potassium hydroxide and treating with chlorine. The 

 weighed oxide was of a salmon color, and when dissolved as 

 nitrate its solution showed no absorption bands. Didymium 

 and lanthanum could be detected only in traces. In the sol- 

 uble yttrium group, the nitrate solution showed the erbium 

 bands quite markedly. Fluorine, zirconium, beryllium, ura- 

 nium, and lithium were carefully tested for, but found to be 

 absent. The basic portion of this mineral is largely composed 

 of the cerium and yttrium metals, while silica and the oxide 

 of titanium appear to the extent of nearly 6 per cent. From 

 the careful manner with which, the mineral used for analysis 

 was selected and purified and from the microscopic character 

 of the fragments used, it appears as though the silica and 

 titanium oxide were normal constituents of the mineral. 



Fluorine could not be detected by even the most careful 

 examination. The high content of water, 5*95 per cent, given 

 off above 110 degrees and the absence of fluorine may be due 

 to the fact that fluorine and hydro xyl are mutually interchang- 

 able in this group of minerals, as they are in the lithia fluorine 

 micas, and in triploidite and related minerals of the wagnerite. 

 group. The isomorphism between hydroxyl and fluorine was 

 first demonstrated by the late Prof. Penfield, who showed its 

 existence for the first time in triploidite, and later in many other 

 minerals in which these radicals occur. 



The analyses of marignaeite and the several members of the 

 pyrochlore group given in the table show the presence of 

 either or both of these radicals in each of the completed 

 analyses. While a careful determination of the hydroxyl in the 

 analyses of these minerals has apparently not been carried out, 

 it seems reasonable to apply the principle of isomorphism of 

 these radicals in this group. 



The presence of an appreciable content of silica, 3*10 per 

 cent, in the pyrochlore from Wausau, perhaps needs further 



* American Chemical Society, xxvi, 1235. 



f Ibid, xxvii, 1369. 



% This Journal, xxii, 520, 1906. 



