CHEMISTRY. 



113 



ble portion of the superphosphate. In order 

 to prevent the precipitation of calcium car- 

 bonate from any sulphate of calcium that may 

 be present, the entire amount of calcium is 

 precipitated by sodium oxalate, and the acidity 

 of the resulting mono-sodium phosphate is de- 

 termined. If free sulphuric or phosphoric acid 

 is present in the superphosphate, sodium car- 

 bonate is added before titration. until the 

 liquid becomes slightly turbid. 



Mr. L. Marquardt has described a new meth- 

 od for the quantitative determination of fusel- 

 oil in brandy. The oil is extracted with chlo- 

 roform, and the product is oxidized with bi- 

 chromate of potash, distilled, and treated with 

 barium carbonate. The chloroform and the 

 excess of barium carbonate are removed, when 

 the baryta and the barium chloride are deter- 

 mined by means of nitric acid. The quantity 

 of amylic alcohol or fusel-oil is calculated from 

 the baryta. 



G. Larsen has shown that copper and zinc 

 can be separated by one precipitation with hy- 

 drogen sulphide if hydrochloric acid is added 

 to the hydrogen sulphide with which the pre- 

 cipitate is washed. Emil Berglund finds that 

 the method holds good when the amount of 

 hydrochloric acid added to the wash- water is 

 smaller than recommended by Larsen, and fur- 

 ther, that zinc is not precipitated with copper 

 if the amount of hydrogen sulphide in the wash- 

 water is small. 



Spring has communicated the results of ex- 

 periments in producing sulphides by exposing 

 various metals mixed with sulphur, both finely 

 divided, to a pressure of 6,500 atmospheres. 

 Magnesium after six pressings, each time being 

 reduced to filings, gave a gray homogeneous 

 mass having a weak metallic luster. Zinc, after 

 three pressings, gave a sulphide resembling the 

 natural blende ; iron, after four pressings, gave 

 a block which was hardly touched by the file, 

 and appeared homogeneous under the micro- 

 scope. Cadmium sulphide was formed easily 

 in three pressings, in a yellowish-gray, homo- 

 geneous mass. Bismuth sulphide and antimo- 

 ny sulphide were formed in two pressings, and 

 lead sulphide still more easily ; copper and tin 

 yielded the sulphide in th-ee pressings. Silver 

 required from six to eight pressings before a 

 homogeneous mass could be obtained. Alumi- 

 num and carbon gave imperfect results. Spring 

 has drawn the conclusion from his experiments 

 that allotropic states are only different con- 

 ditions of polymerization, in which the chemi- 

 cal activity decreases as the process goes on. 



The method almost exclusively employed for 

 estimating the halogens in organic compounds, 

 that of Carius, consists in heating the substance 

 in a sealed tube with fuming nitric acid and 

 silver nitrate. R. T. Plimpton and E. E. Graves 

 propose, in the case of volatile compounds, a 

 method by which the substance is introduced 

 into a U-tube through which illuminating gas 

 and air are passed, as in a Bunsen burner. The 

 volatile substance evaporates and burns with 

 VOL. xxin. 8 A 



the gas at a tip on one end of the tube, while 

 the halogens are left partly tree and partly in 

 combination with hydrogen. The evaporation 

 is sometimes aided by warming the tube with 

 hot water or otherwise. The temperature is 

 raised or lowered so that the substance may 

 always be detected in the flame, yet not in 

 sufficient quantity to make the flame luminous. 

 The products of combustion are aspirated 

 through a bent funnel tube and collected in 

 dilute caustic soda ; this is boiled with sulphur- 

 ous acid to reduce chlorates, etc., and the halo- 

 gens are then precipitated with silver nitrate. 

 The success of the experiment depends upon 

 the regular volatilization of the compound. 



Mr. W. G. Strype, of Wicklow, Ireland, has 

 devised a method of purifying the hydrochloric 

 acid used in the manufacture of chlorine from 

 sulphuric acid before admitting it into the 

 chlorine-stills, by which the difficulties arising 

 in this manufacture from the accumulation 

 of calcium sulphate are to a large extent ob- 

 viated. His process depends upon the fact that 

 while calcium sulphate is somewhat freely 

 soluble in hot hydrochloric acid, it is only 

 slightly soluble in the cold acid. 



Successful experiments have been made by 

 M. J. Garnier at works near Rouen, France, 

 with a new process for removing arsenic and 

 antimony from copper. It comprises the em- 

 ployment of a sole of chalk and tar, over which, 

 for each separate operation, is placed a falsa 

 sole of limestone and manganese peroxide. 

 With the melting of the copper, a generation 

 of carbonic acid and oxygen begins from the 

 upper sole, which oxidizes the charge. As 

 soon as the metal is sufficiently liquid, the 

 lime and manganese ' protoxide rise and dis- 

 solve the arsenic acid. By this one operation 

 the amount of arsenic, according to M. Gar- 

 nier, is reduced to one-fifth. Subsequent fu- 

 sions with basic fluxes are said almost com- 

 pletely to eliminate the arsenic. 



Dr. Sidersky bases a method for the separa- 

 tion of calcium from strontium on the state- 

 ment that on adding a mixture of sulphate and 

 oxalate of ammonium to a solution of stron- 

 tium, the latter is all precipitated as sulphate ; 

 while, if the mixture is added to a calcium salt, 

 only oxalate is precipitated. If it is added to a 

 solution containing both strontium and calcium, 

 the former is precipitated as sulphate and the 

 latter as oxalate. The two precipitates are 

 separated by the solubility of the oxalate in 

 acids. 



Otto F. von der Pfordten has published a 

 new method for the estimation of tungstic acid 

 .by reduction with zinc and hydrochloric acid 

 to tungsten dioxide. The reduction is best 

 effected by using a 27-per-cent. solution of hy- 

 drochloric acid. The solution first becomes 

 blue, then a black-green, and finally a dark 

 brownish-red, the end product being tungsten 

 dioxide, which is determined by titration with 

 potassium permanganate. The method has the 

 disadvantage that only very small quantities of 



