Chemistry and Physics. 505 



mass of the gas contained in the globe used was 2-30143, 2-29890, 

 2-29816, 2-30182; the mean being 2-30008. Two experiments on 

 nitrogen prepared by reducing N 2 by iron, gave 2-29869, 

 2-29940; the mean being 2*29904. Two experiments on nitrogen 

 from ammonium nitrate, passed over hot iron, gave 2*29849, 

 2-29889 ; the mean being 2*29869. If now, we compare with 

 these numbers, the values obtained from nitrogen prepared from 

 the atmosphere, we find that in four experiments, the mass of the 

 gas in the globe was 2-31017, 2-30986, 2-31003, 2-31007 ; the mean 

 being 2-31003. Moreover, on removing the oxygen by placing 

 air in a vessel containing a mixture of slaked lime and ferrous 

 sulphate, the nitrogen gave the values 2-31024, 2-31010, 2-31028; 

 mean 2-31020. Thus the result obtained from air in 1892 with 

 hot copper 2-31026, that of 1893 by hot iron 2-31003, and that of 

 1894 by ferrous sulphate 2-31020 agree well together. It would, 

 therefore, seem that the nitrogen prepared by the four chemical 

 methods given is materially lighter than the nitrogen prepared 

 from atmospheric air, the difference amounting to about 11 milli- 

 grams on the mass contained in the globe, or about one two-hun- 

 dredth part of the whole. Moreover the lighter nitrogen did not 

 become denser by standing for eight months. It was this anom- 

 aly in density that suggested to Rayleigh and Ramsay the pos- 

 sible presence in the air of some new constituent.* — JProc. Roy. 

 Soc, April, 1894 ; Nature, 1, 157, June, 1894. g. f. b. 



3. On Carbon Boride. — It has been shown by Moissan that 

 when the electric arc is allowed to pass between two carbon elec- 

 trodes agglomerated by means of a solution of boric acid and 

 aluminum silicate, or when boron itself is placed in the electric 

 arc, the carbon and boron unite to form carbon boride. It may 

 be obtained in larger quantity by heating a mixture of 66 parts 

 of amorphous boron and 12 parts of carbon made from sugar, in 

 an electric furnace for six or seven minutes, by means of a cur- 

 rent of 200-300 amperes and 70 volts. A black graphitoidal 

 mass is obtained having a brilliant f racture. By treating this, 

 fh-st with fuming nitric acid and then with a mixture of nitric acid 

 and potassium chlorate, the product is obtained in the form of a 

 crystalline powder. Carbon boride may also be prepared by add- 

 ing boron and carbon in excess to iron, and heating the whole in 

 the electric furnace. Better results are obtained however by 

 using copper or silver instead of iron ; the crystals being better 

 defined with the latter metal. Prepared in any of these ways 

 carbon boride has the composition CB 6 and forms brilliant black 

 crystals of sp. gr. 2-51. Chlorine attacks it below 1000°, but it 

 is not affected by bromine, iodine or sulphur, at the softening- 

 point of glass, or by phosphorus or nitrogen at 1200°. Boiling- 

 acids do not act on it, nor do chromic acid, fuming nitric acid or 

 concentrated iodic acid when heated with it in a sealed tube to 

 158°. At 500° no carbon dioxide is formed when the boride is 

 heated in oxygen; but at 1000° it burns slowly and with more 



* This Journal, III, xlviii, 34.5, October, 1894. 



