Oct. 24, 1878] 



NATURE 



679 



hypothesis ought to conrey to him the idea of blue, the 

 error, although not appearing worse to the normal-eyed 

 than calling grass red, would be altogether inconsistent 

 with the proper colour-blind sensations, and would prove 

 that such a person had not dichromic vision in the sense 

 here intended.' 



It may further be remarked that, as the colour-blind 

 person finds the yellow sensation much predominating in 

 what he sees, and as he will have observed that it 

 corresponds to a larger number of ordinary colour-words 

 than the blue sensation, his vocabulary for this group will 

 be naturally more copious than for the opposite one. 



William Pole 

 (To be continued^ 



CHEMICAL NOTES 



Estimation of Mercury.— Prof. Clark, of Cincin- 

 nati, describes {Ber. d. deiitsch. chem. Gesell. xi. 1,409) 

 an interesting application of electrolytic decomposition 

 for the purposes of analytical chemistry, viz., in the 

 separation and estimation of mercury. The solution of 

 the mercury salt, acidulated with sulphuric acid, is placed 

 in a platinum vessel, connected with the zinc pole of a 

 Bunsen bichromate battery of six cells. A piece of 

 platinum foil in connection with the carbon pole is 

 dipped into the liquid, and the decomposition com- 

 mences at once. At first a mercurous salt is precipi- 

 tated. This is reduced gradually, until, in the course of 

 an hour, it is completely changed into the metal, which 

 requires simply to be separated from the solution, washed, 

 dried, and weighed. Similar methods for the estimation 

 of zinc, nickel, and copper have for some years been in 

 use. 



Separation of Antimony and Arsenic. — One of 

 the problems in analytical chemistry awaiting solution is 

 a satisfactory separation of arsenic from antimony. In 

 the last number of Liebig's Annalen (vol. 192) Prof. 

 Bunsen presents a new method intended to supersede 

 that hitherto employed, which was discovered by him a 

 number of years since, depending on the treatment of 

 the combined sulphides with sulphurous acid. In the 

 new process the sulphides are dissolved in potash and 

 subjected to the action of chlorine. A quantity of a 

 saturated aqueous solution of sulphuretted hydrogen, 

 sufficient to precipitate the antimony, as SbjSg, is then 

 added, and in the filtrate the arsenic acid is precipitated 

 on heating as AsjSg by a long-continued stream of HaS. 



Specific Heats of Mercury and Iron.— O. Pet- 

 tersson and E. Hedelius have recently made careful 

 determinations of the specific heat of mercury and iron 

 in the following manner {Ofvers. f. Vetensk. ForhandL, 

 37, P- 35): — Apiece of wrought iron was heated in an 

 air bath to 26°, and then plunged in baths containing 

 weighed quantities of mercury ^^and water at 0°. The 

 resultant temperatures gave the specific heat of iron as 

 referred, firstly to water, and secondly to mercury, and 

 the division of the first value by the second yielded the 

 specific heat of mercury referred to water. The averages 

 derived from twenty experiments give for the average 

 specific heat of wrought iron between 4° and 27°, o" 10808; 

 and for the specific heat of mercury between 0° and 5°, 

 0-033266. The authors find that the specific heat of 

 mercury suffers but slight alterations between o°and 100°. 



Latent Heat of Water at Temperatures below 

 0° C— O. Pettersson {Ofvers. f. Vetensk. Forhandl., 2>7, 

 p. 53) has lately determined the latent heat of water at 

 temperatures below 0°. For this purpose thin tubes con- 

 taining water were placed in a mercury calorimeter, 



' Some errors might, however, legitimately arise in the use of the words 

 for red and green, from the fact that some hues of these colours give yellow 

 sensations, while others give blue sensations. 



cooled to certain temperatures below 0°, and congelation, 

 was induced by the insertion of a snow crystal. The 

 latent heat of water at 0° according to Regnault is 79 '25. 

 The results obtained by Pettersson at lower temperatures 

 are as follows : - 2-80°, 7771 ; - 4'995°, 76"6o ; - 6*28°^ 

 75-94 ; - 6-50°, 76-03 ; - 6-62°, 75-99 ; all of them coin- 

 ciding closely with the estimations of the theoretical 



formula —— = c - h, where r represents the latent heat 

 o 1 



of fusion, T the absolute temperature, and c — h the dif- 

 ference between the specific heats of the solid and liquid 

 body. Experiments were likewise made with sea water 

 containing 3-536 per cent, of solid matter, and freezing at 

 - 9°. At this temperature pure water would possess a 

 latent heat of 75 ; the sea water possessed on the con- 

 trary but 54, showing that the above proportion of saline 

 matter was sufficient to cause a diminution of 28 per cent, 

 in the latent heat. 



Preparation of Salts of Nitrous Oxide.— In the 

 Journal of the Chemical Society (clxxxix.) Mr. A. E. 

 Menkd describes some of the above salts. In analysing 

 a sample of cast iron an experiment was made attempt- 

 ing the conversion of the phosphorus contained in it by 

 fusion with nitre and sodium carbonate, into an alkaline 

 phosphate. During the operation a bulky yellow precipi- 

 tate was obtained which proved to be, not a phosphate,, 

 but identical with the body obtained by Dr. Divers in the 

 action of sodium amalgam on sodium nitrate. The 

 analysis of the silver salt gave a mean percentage of 

 78 09 Ag, agreeing therefore with the formula AgNO, 

 which requires 78-26 per cent. Ag. The salt may also be 

 obtained by the simple fusion of iron filings with nitre, 

 the best heat to employ being that of a charcoal furnace. 

 The analysis of the sodium salt obtained by the fusion of 

 iron filings with sodium nitrate gave numbers closely 

 agreeing with the formula NaNO -^-■ 3H2O. The substi- 

 tution of zinc for iron filings failed to produce the body. 

 On acting on the silver salt with ethyl iodide the silver is 

 converted into iodide, and on fractionating the distillate 

 evidence of the formation of an ethereal salt of low boiling- 

 point is obtained. 



On Manganese Tetrachloride (MnCl4). — Some 

 doubt still existing with regard to the decomposition of 

 manganese oxides higher than the dioxide Mn02, M^* 

 W. Fisher has recently made experiments bearing upon 

 this point. The oxides employed are the sesquioxide,. 

 MngOg, and the red oxide of manganese, Mn304. The 

 analyses of the liquids obtained by the action of the acid 

 on the different oxides were made by decomposing the 

 freshly-prepared solutions with potassium iodide, and 

 then titrating the amount of iodine liberated in each case 

 with sodium hyposulphite. From his experiments Mr. 

 Fisher finds that the higher oxides when treated with 

 excess of acid give a brown liquid containing a highly- 

 chlorinated manganese compound easily resolved into 

 manganous chloride and free chlorine, and on dilution 

 with water yielding manganese binoxide in both instances. 

 The solutions appear to be identical, probably containing 

 MnCl4 in each. Under the conditions of the experiment 

 the corresponding chlorides, MngClg and Mn^Cls, do not 

 appear to be formed from their corresponding oxides, nor 

 do they appear as products of the partial dechlorination 

 of the tetrachloride. The action of the acid on the two 

 oxides the author considers may be represented by the 

 formulae — 



MtijOg + 6HC1 = MnCl.^ + MnCli + 3H2O, 

 Mn304 + 8HC1 = 2MnCl2 + MnCl4 + 4H2O, 

 and as a large excess of acid or alkaline chloride renders 

 MnCl4 more stable, he thinks it probable that this body- 

 may exist in a form analogous to chloroplatinic acid. 



Spontaneous Ignition of Hydrogen by finely- 

 divided Zinc. — In dissolving zinc in hydrochloric acid 

 P. W. Hofmann has observed explosions on the surface 



