March 6, 1902 J 



NA TURE 



42 [ 



facture. The electrodes are first made up in amorphous carbon, 

 which is then converted to graphite by baking at a very high 

 temperature in the electric furnace. It is claimed that the con- 

 version to graphite is complete. These electrodes are especially 

 suitable for electrolytic purposes on account of their high con- 

 ductivity and low porosity, and also, a recommendation which 

 will appeal to experimentalists, on account of the ease with 

 which they can be machined and worked. It is also not unlikely 

 that the use of graphite electrodes in metallurgical processes will 

 extend. Both in metallurgical and electrochemical work the 

 quality of the electrodes is of the first -.importance, making in 

 niiny instances all the difference between success and failure. 



Dr. C. Jensen, of Hamburg, has contributed to the 

 Meleorohgische Zeilschrift for December last an interesting paper 

 on the subject of atmospheric polarisation. The author reviews 

 all the principal theories and discoveries from.those of Arago in 

 1809 to the recent important works of Dr. Pernter relating to 

 the analogy of various turbid media with the blue light of the 

 sky. The author considers that the observation of atmospheric 

 polarisation should be of use in weather prediction, as showing 

 approaching cloudiness some hours before it is visible by other 

 means, and also as showing whether the sky is clear or otherwise 

 above a stratum of fog. 



The " campylograph " invented by Pere Marc Dechevrens for 

 mechanically describing certain ornamental curves is the subject 

 of a paper by its inventor in Cosmos for February 22, and of a 

 pamphlet, published in Brussels, by Pere Potron, who deals with 

 the equations of the curves it describes. The machine, to which 

 the name campylograph is given, is an arrangement for com- 

 pounding the projections, in two mutually perpendicular direc- 

 tions, of two circular motions, and as such it can be made to 

 trace Lissajou's curves exactly, the amplitude of the component 

 vibrations remaining constant instead of decaying as is the case 

 with compound pendulums. Moreover, the table carrying the 

 paper can be rotated and a variety of figures thus obtained, 

 including the epicycloids and hypocycloids, and also curves 

 similar to those given by a harmonograph with clockwork table, 

 but without the gradual decrease in amplitude. It should, how- 

 ever, be remembered that numerous machines have at different 

 times been designed for drawing various classes of ornamental 

 curves, and it seems as likely as not that something practically 

 identical with the ' ' campylograph " may have been previously 

 constructed. 



Several papers tending to raise doubts regarding the well- 

 known laws of electromagnetism for bodies in motion have 

 already been noticed in these columns. M. R. Blondlot writes 

 on this matter in the Journal de Physique for January. If a 

 current of air is moving parallel to the axis of 2 in a magnetic 

 field the lines of force of which are parallel to the axis of x, then, 

 according to Hertz and Lorentz, an electromotive force ought 

 to be set up in the negative direction of the axis of y, and if the 

 two plates of a condenser are placed perpendicularly to this 

 direction and connected by a wire so as to bring them to the 

 same potential, they ought to become charged and to remain 

 charged when they are separated. According to M. Blondlot, 

 however, the only observable effects are such as can be ac- 

 counted for by accidental causes, and are small in comparison 

 with those required by theory. The author further remarks 

 that by an application of the principle of action and reaction it 

 would appear to follow that a displacement current in air exerts 

 no magnetic action, and according to this view the discharge of 

 a condenser is magnetically an open current. Or if this contra- 

 diction of Maxwell's theory be not admitted, we should have to 

 abandon the principle of reaction. 



NO. 1688, VOL. 65] 



The spark spectra of those elements which are gaseous at 

 ordinary temperatures have been extensively investigated, but 

 little research has been hitherto done on the arc spectra of gases. 

 Mr. O. H. Basquin has recently published, in the Proceedings 

 of the American Academy of Arts and Sciences (vol. xxxvii 

 pp. 161-174), an account of his research on "The Arc Spectrum 

 of Hydrogen." In his apparatus, one electrode was made to 

 rotate quickly, this preventing the welding together of the elec- 

 trodes and also throwing the hot gases to one side. That part 

 of the flame of the arc thus separated from the poles was very 

 free from continuous spectrum, and he was thus better able to 

 study the spectrum lines than if a stationary arc were used. The 

 arc was enclosed in a comparatively air-tight brass "hood," 

 through which a stream of electrolytically-prepared hydrogen 

 was continually passing. The light from the arc passes through 

 a lens fixed at the end of a brass tube fitting into the wall of the 

 hood. The arc spectra in hydrogen of aluminium, copper, mag- 

 nesium, coin-silver, sodium, tin and zinc were examined both 

 visually and photographically. Of the well-known series of 

 hydrogen lines seen in its low-pressure spark spectrum and in the 

 spectra of the hotter stars, Ha and H;8 are well shown visually in 

 the case of all metals except Na ; the majority also show H^, but 

 II j is rarely seen. Ha is sharp and well defined, the others 

 broad and hazy. In the photographic spectra, II3 and Hy are 

 seen in the case of all metals except Na, while lis shows in the 

 spectra of tin, silver and copper. A small dispersion spectrum 

 of tin shows He also. The weaker hydrogen lines more re- 

 frangible than He have not been traced, this being probably 

 due to the inordinate width of the lines, those detected averag- 

 ing about twenty-five tenth metres. All the metals except tin 

 give a characteristic set of lines which do not occur in the arc 

 spectrum in air of the corresponding metal. These new lines 

 do not seem to bear any particular relation to the spark lines of 

 the respective metals, and the author supposes them to be due to 

 compounds of hydrogen with.the metals formed in the arc. No 

 new lines have been found which can with certainty be attributed 

 to hydrogen. At the end of the paper an interesting discussion 

 is given of the general effects of the hydrogen atmosphere on the 

 arc and its spectrum. 



The first of the international balloon ascents for the curren' 

 year took place on January 9. At Chalais-Meudon the ascent 

 was at 8h. a.m.; temperature at starting i°,C., maximum height 

 reached 11,405 m., minimum temperature -63°!. .At Trappes 

 two ascents were made; the heights reached were 15,000 m., 

 temperature -6i°'4, and 15,670 m., -62°'2. At Strassburg a 

 height of 8100 m. was reached ; temperature at starting -4°.4, 

 minimum - 42°-8. One of the manned balloons at Berlin 

 remained up for nearly twenty-nine hours, and the following 

 readings were taken: — at starting, 3°9 ; at 3490 m., -4°'3; 

 at 4850 m., - I5°'i. While travelling eastward the same 

 stratum of cloud was always observed, the upper edge of which 

 continually increased in height. Above the cloud an inversion 

 of temperature occurred and a sudden change of wind from west 

 to about north-west. Three balloons ascended from Vienna ; 

 in one of the manned balloons a temperature of - 10° was 

 recorded at 4100 m. Ascents were also made from Pavlovsk 

 (St. Petersburg) and Mr. Rotch's observatory, Blue Hill, in 

 the United States. Except at St. Petersburg, the balloons 

 ascended in an extensive area of high barometric pressure, the 

 centre of which lay over the Alps. 



In Symons's Meteorological Magazine for February, a map is 

 given showing the places, so far as at present ascertained, at 

 which deposits of yellowish-pink dust were observed on January 

 22 and 23. It attracted attention over practically the whole of 

 Cornwall, near the western border of Devon, and at a few 

 points in Somerset, the south of Gloucester and Glamorgan. 



