-1 A Mi Marignac on the Tungstates. 



represented by the formula MO WO 3 ; but there are numerous 

 acid salts with various and complex constitutions, although all 

 containing 21 to 2£ equivalents of acid for one of base. These 

 are probably complex compounds of bitungstates and tritung- 

 states. 



Among these various salts, which are decomposed by frequent 

 solutions and recrystallizations, there is one class distinguished 

 by their great stability, and which Laurent* called the paratung- 

 stntes. He believed the ratio of the base to acid was as 5 : 12. 

 More recently Lotz was led to propose the ratio 3 : 7, which 

 view Scheibler adopted. Marignac' s analyses have led him to 

 support Laurent's views, or rather to consider them as double 

 salts, and, allowing for the water in their composition, formulate 

 them thus : 



3(MO, HO, 2 WO 3 ) + 2(MO, 2HO, 3W0 3 ). 



Marignac, like Berzelius, has not been able to obtain true 

 fluotungstates : the salts obtained by treating tungstates with 

 hydrofluoric acid contain oxygen, and may be considered as 

 compounds of tungstates and fluotungstates, and represented by 

 the general formula 



MW0 2 F 2 , or M0W0 3 +MPWF: 



In the study of the crystalline form of the fluoxytungstates, 

 one curious result has been noted, namely, that the tiuoxy- 

 tungstate of copper is isomorphous with the group containing 

 the fluosilicate, fluostannate, and fluotitanate of the same metal. 

 With the ordinary formulae, 



Cu F Ti F 2 and CuO WO 3 + CuF W F 3 , 



this isomorphism appears at first sight inexplicable. But this 

 anomaly disappears when it is considered that isomorphism only 

 obtains in formulas when these represent the atomic composition 

 of a body, and not their equivalent constitution. Now fluorine 

 being a biatomic element, it must be doubled, and thus the for- 

 mulas of these compounds become 



CuTiF 6 andCuW0 2 F 4 . 



This curious isomorphism justifies to a certain point the con- 

 clusions, that fluorine can in certain cases replace oxygen atom 

 for atom, spite of the non-equivalence of their atoms; and also 

 that Berzelius's mode of regarding the fluoxytungstates, or salts of 

 tungstates and fluotungstates, cannot be retained, for the isomor- 

 phism of these salts with the fluotitanates cannot be thereby 

 explained. 



The silicotungstates are a new class of salts, but must have 

 been obtained by most chemists who have worked at the tung- 



