Dec. 22, 1887] 



NA TURE 



179 



eyes than his : so perhaps it may be permitted to be said that it 

 was seen in the parish of Llanefydd, Denbighshire, by a corre- 

 spondent of mine, who writes : — " On Tuesday night, November 

 15, while returning homewards on foot, happening to look east- 

 wards I saw a long train of brilliant light suddenly flash out of 

 the sky. At first I thought it was lightning. But instead of 

 vanishing it descended with great rapidity, the light increasing in 

 brilliancy as it neared the earth. The night was rather dark, 

 although the sky was thickly studded with stars, but in a few 

 seconds so intensely brilliant had the light become that a pin 

 might have been picked up from the road with the greatest ease. 

 While I was looking, the object that accompanied the flash burst, 

 and displayed a magnificent mauve and red fringe of light. I 

 say fringe, as it would be impossible for me to describe otherwise 

 the shape, for it appeared to me to project shafts of light, some 

 long and some short, like what would be the rays of a great star. 

 There was in the direction in which I was looking a thick wood, 

 and the effect on the trees of the silvery light I first noticed was 

 richly beautiful. But the effect of the mauve and red light was 

 magnificently grand, and to me in no little degree awful. The 

 whole wood was enveloped in a red lurid glare, svhich lasted as 

 near as I can calculate some six or eight seconds. The effect 

 altogether was like a brilliant transformation scene, and the 

 meteor having passed away, the darkness of the night seemed 

 to be in the last degree intense." 



J. Lloyd Bozward. 



A Correction. 



In the footnote, p. 152, second column, there is an unfortunate 

 transposition for which myself more than the printers are to 

 blame, which i' is important to correct. 



After the words " by Aristotle, Probl. i E 3 and Metaph. 

 A 5 " occur the words " which he attributes to Pythagoras .... 

 on the nature of the Beast." 



These words should have come at the end of the subsequent 

 paragraph where I say that " Muhamad-al-Sharastani assigns 

 reasons for regarding all the numbers up to 10 inclusive as 

 perfect numbers." 



It is these reasons which I speak of a; being by him attri- 

 buted to Pythagoras, &c. 



I may take this opportunity of giving, as another example of 

 the use of the New Nomenclature, the well-known extended 

 Theorem of Fermat, which may be expressed by saying : 



'•' Every number must divide the Fermatian of which the index 

 is its tot lent, and the base any one of its iotitives." 



Athenpeum Club, December 15. J. J. Sylvester. 



ISOLATION OF FLUORINE. 



ONE of the most difficult problems of modern chemis- 

 try has at last been satisfactorily solved. After 

 three years of incessant labour, occasionally interrupted 

 by temporary feelings akin to despair, M. Henri Moissan 

 has at length isolated in considerable quantities that most 

 baffling of elements— fluorine, and has been enabled to 

 determine its principal properties. The experiments them- 

 selves are among the most interesting ever performed, 

 and their details, as described by M. Moissan in the 

 December number of the Ajinales dc Chimie et de Phy- 

 sique, form the most fascinating reading. They must of 

 necessity have been extremely costly, for by far the 

 greater portion of the apparatus employed was constructed 

 of platinum, and it is not often that one hears of a plati- 

 num tube 80 centimetres long and of i^ centimetre 

 diameter being destroyed in each experiment, as happened 

 in the earlier stages of these researches. 



The isolation of fluorine has formed a worthy object of 

 the attention of chemists ever since the first remarkable 

 experiments of Sir Humphry Davy, who was rendered 

 dangerously ill by being exposed to the corrosive fumes 

 of hydrofluoric acid. Although Davy was not successful 

 in obtaining free fluorine, yet he brought clearly to light 

 the nature of hydrofluoric acid, and proved it to consist 

 of hydrogen combined with an unknown but extremely 

 active element— fluorine. The history of all the attempts 

 which have since been made to effect the preparation of 



free fluorine might occupy a volume, and it will therefore 

 only be necessary to refer to the later work of our 

 countryman. Gore, who, in 1869, published his researches 

 upon the electrolysis of hydrofluoric acid, and of certain 

 fluorides, and left our knowledge of the acid itself in a 

 most complete state. M. Moissan, working in the labor- 

 atory of M. Debray, now steps in and achieves the result 

 so ardently sought after during the last eighty years — 

 another example of the irresistible power of human per- 

 severance. 



In the light of the experience gained by former experi- 

 menters, it appeared that the action of a powerful electric 

 current upon the compounds of fluorine with the non- 

 metallic elements, such as hydrogen, phosphorus, and 

 arsenic, would be most likely to yield the desired result ; 

 knowing also that fluorine must be an extremely energetic 

 substance, it was absolutely essential to work at very low 

 temperatures. Hence M. Moissan's first attack was made 

 upon the fluorides of phosphorus and arsenic, but finding 

 these to be practically impregnable, he diverted his attack, 

 guided by certain indications afforded during his first 

 attempt, upon hydrofluoric acid itself. Finding, however, 

 that pure hydrofluoric acid is an exceptionally bad con- 

 ductor of electricity, as has been stated by other workers 

 — that even a current from fifty Bunsen cells would not 

 pass through the liquid — he eventually, after several essays, 

 succeeded in converting it into a conductor by dissolving 

 in it a quantity of the double fluoride of potassium and 

 hydrogen. On passing the current from twenty Bunsen 

 cells through the now conducting medium, hydrogen im- 

 mediately commenced to be evolved at the negative 

 terminal, while fluorine was with similar rapidity evolved 

 at the positive pole, and exhibited its tremendous activity 

 upon everything that came near it : burning up hard 

 crystalline silicon like tinder, setting fire to organic 

 matter, and forming fluorides with incandescence with 

 many other elements. 



Having thus indicated the general course of these 

 researches, it will no doubt be interesting to follow M. 

 Moissan during the carrying out of his principal experi- 

 ments. 



The first series consisted in examining the action of 

 electric induction sparks upon the gaseous fluorides of 

 silicon, phosphorus, and arsenic. The gases were intro- 

 duced into glass eudiometer tubes standing over mercury, 

 and the spark was passed between two platinum wires 

 connected with an induction-coil actuated by a few 

 Grenet or Bunsen cells. On introducing dry silicon tetra- 

 fluoride, SiF4, and passing sparks for an hour, no decom- 

 position was effected, the result being discouragingly nil. 

 Dry phosphorus trifluoride, PF3, however, behaved quite 

 differently, phosphorus being deposited upon the inner 

 wall of the tube ; but the fluorine liberated at once com- 

 bined with the residual trifluoride to form the more stable 

 pentafluoride, PF5. Some time ago this pentafluoride of 

 phosphorus was prepared by Prof. Thorpe, who also sub- 

 mitted it to the action of the induction-spark, unfortun- 

 ately without effecting any decomposition. Precisely the 

 same result has been arrived at by M. Moissan, using a 

 004m. spark ; but on obtaining sparks o'2m. long, a rapid 

 etching of the walls of the glass tube occurred, and the 

 meniscus of mercury entirely lost its brilliancy. After an 

 hour's duration the experiment was concluded, and the 

 apparatus allowed to cool, when it was noticed that the 

 volume had diminished ; moreover, the gas was found to 

 have changed its properties, yielding a precipitate of 

 silica in contact with water, while the residual gas con- 

 sisted of the trifluoride of phosphorus. Hence PFg = 

 PF3 -f Fo, which latter forms, with the glass, silicon tetra- 

 fluoride, and, with the mercury, fluoride of mercury. So 

 here again the experiment was disappointing, and al- 

 though fluorine was for the moment liberated, this method 

 was "certainly not suitable for the preparation of free 

 fluorine. 



