Intelligence and Miscellaneous Articles. 315 



called basic salt. Of the methods recommended for analysing an 

 aluminous precipitate, especially after it has been heated to redness, 

 the author found that of Otto, by means of a magnesian salt, the best. 

 It is not however applicable to quantitative analysis, as it always 

 yields too little. 



l;.'t; i nil iliilU „. . , -' 



, 'li ii'Mf?/' Viviamte. 



We have already several analyses of this mineral, which however 

 differ considerably, and do not give its real composition, as all the 

 iron it contains is estimated as protoxide, although both the protoUff 

 and peroxide enter into its composition. They are as follows : — ii- 



Fromthelsleof Bodenmais From St. Agnes in Prom New Jersey. From Kertsch tf' 

 France (Laugier). (Vogel). Cornwall (Stromps^). ,. ,('i;hoi5^8ftn), ^ , Crimea (Segetl^yi 

 Phosphoric acid 21 26-4 31^825 _, ilgg, ,. 24-95 



Protoxide of iron 45 41-0 41-2266 '^^^'' ^^^'^B-dl""^ ' 48-79 



Water ._34 31;0 27-4834 2714 2626 ' 



100 98-4 99-8934 99-51 100* ' ' 



The author analysed a variety of vivianite, which was indistinctly 

 crystallized in the form of needles ; it came from the Mulica hills in 

 New Jersey (and was probably the same as that analysed by Thom- 

 son, and which he called Mullicite); — and also the well known qry- 

 stalline variety from Bodenmais. .. /; 



1. Viviamte from New Jersey. — a. The solution of the mineral in 

 muriatic acid was precipitated by ammonia and hydrosulphuret of 

 ammonia ; the sulphuret of iron was dissolved in nitro-muriatic acid 

 and precipitated by ammonia ; the phosphoric acid was thrown down 

 from the former solution by chloride of calcium. In a second expe- 

 riment (b) the fossil was heated to redness in dry hydrogen gas, and 

 the water which escaped was estimated. 



The muriatic solution was evaporated to dryness, then fused with 

 carbonate of soda, to estimate the oxide of iron. Finally (c), a por- 

 tion of the mineral was boiled with dilute muriatic acid and a weighed 

 portion of copper, until the whole of the iron was reduced to the 

 state of protoxide. 



2. Vivianite from Bodenmais. — The mineral was at first treated as 

 above with copper, the solution freed from copper by sulphuretted 

 hydrogen, evaporated to dryness, the residue heated to redness with 

 carbonate of soda, washed with water, and thus the oxide of iron 

 and phosphoric acid estimated in the usual way. If we first of all 

 suppose the whole of the iron to exist in the form of peroxide, the 

 analyses yield, — 



tt, tU — i^i ■• . '"■•From New Jersey. From Bodenmais. 



„,j.„ *. . <-■ ,', «. b. 



Phosphoric acid. ..... 28-40 29-01 



Peroxide of iron 49-74 49-36 51-21 



Water 27-49 



But the peroxide of iron estimated in c amounts in the vivianite. of 

 New Jersey barely to 12-06 per cent., in that of Bodenmais to 

 11-60 per cent., so that the remainder must be considered as prot- 

 oxide of iron. The mean quantities of these results calculated with 



