148 W. T. ScJialler — Ludwigite from Montana. 



Discussion of JForinula. 



From the titjjures of the average analysis, the ratios were 

 calculated and found to be as follows : 



Ratios of ludwigite analysis. 



M<rO 9760 ) ..„_ . ^. ( 3-57 MgO 



FeO 804 r^32 ^OOJ .43j,»3 



FeA 1858 [Less 1475 M.-^SiO, [ 



A1.A 177 and82MgC03] f^^'^^ ^ "^ 



SiO, 1475 



CO, 82 



B.,03 1926 1926 1-02 



The ratios agree well with the formula generally given for 

 ludwigite, 4E,O.Fe„03.B./^j The high water content obtained 

 by Whitfield,* on the Hungarian material, namely 3*62 per 

 cent, is substantiated neither by the analysis of the Montana 

 mineral, as given above, nor by the analysis of the material 

 from Hungary, as described beyond. Yet both samples yield 

 about a per cent of water which is at present impossible to 

 prove foreign to the mineral and may, in time, be shown to be 

 an essential constituent thereof. If the water given off above 

 107° (HjO+) be considered as essential to the mineral, a 

 formula of about the proportions RO : Fe^O, : Bfi^ : H„0 = 

 16 : 4 : 4 : 1 would result. 



While the ratios agree well with the genera^l formula of 

 ludwigite, the amount of ferrous iron is too small (even by 

 referring all the ferrous iron in the analysis to ludwigite and 

 none to either the forsterite or the carbonate) for the formula 

 FeO.Fe„03.3MgO.B,03. From the data given above, the for- 

 mula for the Montana ludwigite may be written 'STMgO. 

 •43 FeO. 3MgO. B,0,j, showing an isomorphons mixture of a 

 ferrous ferric magnesian borate with a magnesian ferric 

 magnesian borate, the latter being in excess. 



As has been previously shown by the analyses of Ludwig 

 and Sipocz and of Whitfield and confirmed by the analysis 

 given beyond, the Hungarian mineral approximates closely to 

 the definite formula FeO.Fe203.3MgO.B.^O.j, though one anal- 

 ysis by Ludwig and Sipocz shows a slight deficiency in the 

 ferrous iron. The mineral from Montana, however, is very 

 much lower in the ferrous iron and correspondingly higher in 

 the magnesia, the ratios of B^Oj : Fe^O, : FeO-1-MgO remain- 

 ing constant. There is here, then, sufiicient evidence of the 

 existence of magnesian ferric borate, MgO.Fe^Og.SMgO.B^Oj, 

 free from ferrous iron, not yet found in nature in the pure 

 state but preponderant in the Montana ludwigite, being present 

 * Whitfield, this Journal, vol. xxxiv, p. 284, 1887. 



