On the Preparation of Solutions of Peroxide of Hydrogen. 1 i') 



sten, and the iron and manganese are then got rid of by dilute 

 hydrochloric acid. 



The author also succeeded in proving the existence in this 

 mineral of tantalum, by means of Marignac's test with fluoride 

 of potassium ; but he was unable to obtain tantalic acid quite 

 free from niobic acid. 



The following resume of a memoir on a new mode of testing 

 mineral oils is given by the authors, MM. Salleron and Urbain*. 



We have sought a method of testing the mineral oils employed 

 in illumination which should be more exact than the determina- 

 tion of their density, or the direct measure of their inflammability. 

 We show the inconveniences of each of these methods, and pro- 

 pose to substitute for them the measurement of the tension of 

 the vapour of these liquids, which is proportional to their vola- 

 tility, and therefore to their inflammability. We show that this 

 method is more delicate than former ones. We then give the 

 description of a new apparatus by which these tensions may be 

 determined, and finally a Table containing the elastic forces of 

 the vapour of one and the same oil taken as type ; so that know- 

 ing, on the one hand, the tension of the oil to be tested corre- 

 sponding to a given temperature, and, on the other, the tension 

 at this temperature of the typical oil, by comparing these num- 

 bers the value of the specimen examined can be deduced. From 

 our experiments we conclude that a tension of 64 millims. of 

 water might be adopted as the limit of that which oils furnished 

 to the public should possess. 



Hoffmannf describes the following method of readily obtain- 

 ing a pretty strong solution of peroxide of hydrogen. Potassium 

 is melted in a porcelain crucible, and a current of air is blown on 

 it by means of a bellows ; in this way a yellowish-green mass is 

 obtained which is rich in KO 3 . This is then added in small 

 portions to tolerably concentrated fluosilicic acid, by which a 

 certain quantity of oxygen is disengaged, at the same time that 

 silicofluoride of potassium precipitates. The liquid, which is 

 advantageously left acid by a small excess of silicofluoric acid, is 

 considerably richer in HO 2 than that obtained by a slow oxida- 

 tion of the metal. Tested with solution of permanganate of 

 potash, the proportion of HO 2 in the liquid amounted to 1 in 

 735. When the silicofluoric acid was kept quite cool by means 

 of a freezing-mixture, this proportion was doubled, and even 

 more ; in one case it amounted to 1 in 357. 



By adding KO 3 to a solution of tartaric acid, a comparatively 

 strong solution of HO 2 may also be obtained. 



* Com,ptes Rendus, January 2, 1866. 

 t Liebig's Annalen, November 1865. 



