;34 



NATURE 



[August 6, 1891 



for the trifluoride of phosphorus and chlorine are found to react 

 in equal volumes, and the combination is attended by a con- 

 traction of one-half. The new gas may therefore be considered 

 as phosphorus chlorofluoride, PCUFg, the chlorine derivative of 

 phosphoryl and thiophosphoryl fluoride, POF3 and P.SF3. 



The most convenient mode of preparation is described as fol- 

 lows. Two flasks of equal capacity (about 500 c.c.) are taken, 

 and filled respectively with phosphorus trifluoride and chlorine. 

 They are connected together by a bent tube passing through the 

 stoppers, and the flask containing the phosphorus trifluoride is 

 further connected with a reservoir of mercury in such a manner 

 that a gentle pressure may be placed upon the trifluoride, .so as 

 to gradually displace it over into the chlorine. The two flasks 

 being of equal capacity, it is evident that, when the whole of 

 the trifluoride has thus been transferred, the reaction is com- 

 pleted, the green colour of the contents of the other flask dis- 

 appears, and the remaining gas is almost pure chlorofluoride. 

 After allowing to stand a few days in contact with the mercury, 

 in order to remove the last traces of chlorine, the gas is ready 

 for examination. 



Phosphorus chlorofluoride is a colourless incombustible gas, 

 possessing a powerfully irritating odour. It is instantly ab- 

 sorbed aid decomposed by water and by solutions of alkaline 

 or alkaline earthy hydrates. A determination of its vapour- 

 density gave the number 5-40, sufficiently near the theoretical 

 density of a substance PCI2F3 (5-46). It is comparatively easily 

 liquefied, a temperature of - 8° C. being sufficient at ordinary 

 pressures. It is dissociated at a temperature of 250° C. into 

 gaseous pentafluoride and solid pentachloride of phosphorus. 

 The induction spark effects the same decomposition. 



Sulphur reacts with phosphorus chlorofluoride in a most in- 

 teresting manner. The reaction commences about the melting- 

 point of sulphur, 1 15° C. , and the products are chloride of sulphur 

 and gaseous thiophosphoryl fluoride, PSF3. And here a most 

 emphatic protest must be made against the manner in which 

 many French chemists persistently ignore the work of the 

 chemists of other countries. Thiophosphoryl fluoride, PSF3, 

 was discovered and prepared three years ago in the Research 

 Laboratory of the Royal College of Science, South Kensington, 

 by Prof. Thorpe and Mr. J. W. Rodger ; and a detailed 

 account, illustrated by experiments, of the mode of preparation 

 and properties of this remarkable gas, was laid before the 

 Chemical Society and published in their Journal. ^ And yet, 

 in the memoir just presented by M. Moissan, we find this 

 compound, a description of which long ago found its way into 

 the abstracts or referate of most foreign journals, described as 

 " un nouveau compose gaseux." Indeed, a considerable amount 

 of unnecessary trouble appears to have been taken in order to 

 ascertain the composition of this " new gas" — trouble which, as 

 the compound is so readily recognizable by its extraordinary 

 properties, might have been saved, if the author had taken the 

 pains to look up the literature of the subject. It is high time 

 that French chemists should look to their "prestige "in this 

 respect, for, unfortunately, the present is by no means the only 

 case which has within the last few months come before the notice 

 of the writer of this note, in which compounds fully described 

 and worked out by English chemists have been rediscovered 

 and described as new by French authors. 



When phosphorus chlorofluoride is passed over free phos- 

 phorus heated to 120°, it is decomposed with formation of 

 phosphorus trifluoride, which passes away as gas, and phos- 

 phorus trichloride, which condenses in liquid drops. Metallic 

 sodium, when slightly heated, appears to absorb the chloro- 

 fluoride entirely, while magnesium, aluminium, iron, nickel, 

 lead, and tin, when heated to about 180°, attack the gas with 

 formation of anhydrous chlorides and liberation of phosphorus 

 trifluoride. Mercury attacks it very slowly at the ordinary 

 temperature, but very rapidly at 180°, with formation likewise 

 of a chloride of the metal and gaseous trifluoride of phosphorus. 

 Hence, when purifying the gas from the last traces of chlorine, 

 the mercury should not be agitated, but allowed to remain at 

 rest, as agitation brings about a perceptible amount of decom- 

 position. 



Water reacts in two stages with phosphorus chlorofluoride. 

 When a little aqueous vapour is admitted into the vessel inclosing 

 the gas, phosphoryl fluoride and hydrochloric acid are formed in 

 accordance with the equation — 



PCI2F3 -f- H2O = POF3 -1- 2HCI. 



When passed into water, however, the gas is completely 

 ' Journ. Chem. Soc. Trans, 1889, vol. Iv. p. 305. 



decomposed into phosphoric, hydrochloric, and hydrofluoric 

 acids — 



PCI2F3 + 4H2O = H3PO4 -1- 2HCI +3HF. 



Ammonia gas reacts at the ordinary temperature with pro- 

 duction of a white solid compound, readily soluble in water, 

 which appears to be fluophosphamide, PF3(NH2).2. 



PC 12^-3 + 4NH3 = PF3(NH2)2 +2NH,C1. 



Phosphorus chlorofluoride is absorbed by absolute alcohol with 

 production of a compound possessing a penetrating odour, and 

 which burns with a bright flime bordered with green, and leaves 

 a white residue of phosphoric acid. The nature of this compound 

 has not yet been fully ascertained. 



These properties of phosphorus chlorofluoride indicate that the 

 gas is much less stable than the pentafluoride, and that the two 

 atoms of chlorine possess a mobility which renders their removal 

 a matter of considerable ease. A. E. Tutton. 



PROF. MENDELEEFF ON THE VARIATION OF 

 THE DENSITY OF WA TER A T DIFFERENT 

 TEMPERA TURES. 

 nPHE last number of the Journal of the Russian Physical and 

 Chemical Society (1891, No. 5) contains an important 

 paper, by Prof. Mendeleeff", upon the variation of the density of 

 water at diff"erent temperatures. In a work, published in 1884 

 and translated into English in the Journal of the Chemi- 

 cal Society, the Russian Professor proposed the formula 

 S^ = So(i - /'/) as a first approximation to a mode of expressing 

 the expansion of liquids at a certain distance from the tempera- 

 tures at which they change their state, and within the limits of 

 accuracy attained in the present determinations. But he re- 

 marked that the expansion of water would require a separate 

 formula, and he now proposes the formula 



S, = i - 



(t - 4)' 



NO. II 36, VOL. 44] 



CA + t){ti - t)C' 

 which embodies, with sufficient accuracy, all that is yet 

 known about changes in the density of water (S^-) within 

 the limits of from - 10" to + 200°. For alU liquids 

 save water, the increase of density with the increase of 



temperature, that is, the derived — , varies but little ; it but 



di 

 slightly increases or slightly decreases with considerable changes 

 ds 

 di 



at + 4°, but very rapidly varies even at temperatures remote 

 from zero, and even superior to 100°. After confirming the 

 above by a few examples. Prof. Mendeleeff" indicates the faint 

 relations between his new formula for water and the general law 

 of the expansion of liquids, by explaining the way in which he 

 arrived at his new formula. He points out, moreover, that 

 under the present state of the determinations of the density of 

 water at various temperatures, it would be impossible to find 

 exact figures for the constants A, B, and C, in the above for- 

 mula, and that provisionally, and especially for temperatures 

 between 0° and 40°, they may be taken as follows : — A = 94'xo, 

 B = 703'5i, and C = 1-90. 



Prof. Mendeleeff then goes on briefly to analyze the various 

 corrections which ought to be taken into account in the deter- 

 minations of the density of water ; namely, the influence of 

 pressure, the expansion of solids, and the measurements of tem- 

 perature. All these being taken into account, it appears that 

 the errors of the best determinations of densities attain several 

 units in the fifth decimals, even at common temperatures. After 

 many unsuccessful attempts at improving the current figures of 

 densities by introducing into them several corrections, Prof. 

 Mendeleeff abandoned the idea, and he now gives the authentic 

 figures, as they were published by the investigators themselves, 

 simply expressing all determinations in volumes for the sake of 

 facilitating comparison. The figures published by Hallstrom 

 {1823), Muncke (1828), Stampfer (1831}, Despretz (1837), 

 Pierre (1847), Kopp (1847), Pliicker and Geissler (1852), Hagen 

 (1855), Henrici (1864), Jolly (1864), Matthissen (1865), Weidner 

 (1866), and Rosetti (1869), are thus given in a first table. The 

 figures, as they were corrected by Biot in i8ii, Hallstrom in 

 1835, Miller in 1856, Rosetti in 1871, Volkmann in i88i, Men- 

 deleeff in 1884, and Makaroff in 1891, are given in a second 

 table. 



The averages of the volumes of water derived from the original 



