564 



NATURE 



{Oct. 9, 1879 



cheeked Capuchin {Cebus lunatui) from South America, presented 

 by Mr. Adrian Hope, F.Z.S. ; an American Red Fox (Cams 

 fnlvus), a Rough-legged Buzzard (Archibuico lagopus) from 

 Labrador, presented by Lord Hobart ; tliree Vulturine Guinea 

 Fowls [Ntimida vuHurina), a Crested Guinea Fowl (Numida 

 cristata) from Last Africa, presented by Vice-Admiral John 

 Corbett, C.B. ; two Malaljar Mynahs (Sturnia malabarica) from 

 Hindostan, a Chinese Mynah {Acridotheres cristatellus) from 

 China, a Waxwing (Amfelis garrulus), European, presented by 

 Mr. A. F. Weiner, F.Z.S. ; five Fat Dormice [Myoxus glis), 

 European, presented by Mr. Edwin Liot; seven Green Tree 

 Frogs (Hyla arborea), a Green Lizard {Lacerta viridis), three 

 Spotted Salamanders [Salamandra maculosa), European, pre- 

 sented by the Rev. S. R. Wilkinson, F.Z.S. ; an Anaconda 

 (Eunectes Jiiurinus) from South America, presented by Capt. E. 

 Ball ; an Elliot's Guinea Fowl (Numida ellioti), a Vulturine 

 Guinea Fowl (Numida vuliurina), three Mitred Guinea Fowl 

 (Numida mitrata) from East Africa, a Booted Eagle (Nisaetus 

 fennatus], European, purchased. 



ON THE GRADUAL CONVERSION OF THE 

 BAND SPECTRUM OF NITROGEN INTO A 

 LINE SPECTRUM 

 pROF. \Vi)LLNER, of Aachen, has recently published a 

 ■*• treatise on the two different views which are held by 

 physicists with regard to the various spectra presented by gases 

 which are rendered incandescent by means of induction sparks. 

 One of these views was first stated by Angstrom, who thought 

 that for a certain gas only one spectrum was possible, and that 

 this spectrum consisted of lines only. All band spectra which 

 occasionally appeared when gases were examined in the way 

 mentioned, he ascribed to impurities. The band-spectrum of 

 nitrogen, according to his idea, belonged to oxides of that 

 element. He believed that as long as the current passed through 

 the gas without giving a spark, the oxide was rendered incan- 

 descent as such, without decomposition, and that the spark 

 decomposed the oxide, and that only then the nitrogen could 

 give its own line-spectrum. Later on Angstrom modified this 

 view, and admitted that an elementary substance might give 

 several spectra when rendered incandescent in the gaseous state, 

 but he still held that in this case the element in question entered 

 into isomeric compounds with itself, and that the different spectra 

 belonged to different isomeric compounds. Mr. Lockyer after- 

 wards defined this view more clearly, stating his opinion that the 

 line-spectrum is produced by simple atoms, and the continuous 

 or channelled-space spectra by conglomerations of molecules. 



Prof. Wiillner, however, does not consider this hypothesis 

 necessary for the explanation of the different spectra of elements, 

 but holds that they may be explained by Kirchhoff's maxim. 

 Prof. Zollner has pointed out that the light emitted by a radiating 

 layer of gas must essentially depend on the thickness and density 

 of the layer. Prof. Wiillner, therefore, after having first con- 

 firmed the fact that the line-spectrum of elementary gases only 

 appears with the real electric spark, the band-spectrum, how- 

 ever, when in the gas the electro-positive brush and glow appears, 

 ascribes the different spectra to the differences in the radiating 

 layers of gas. He believes that in the spark only the molecules 

 struck by the spark are glowing, therefore almost only a linear row 

 of molecules ; thus in the spectrum only the absolute maxima of the 

 emission power, which correspond to the temperature of the 

 spark, become apparent. If, however, in the positive brush 

 light the whole quantity of gas contained in the spectral tube is 

 rendered incandescent, then it is always a relatively thick layer 

 which emits light ; in the spectrum all those kinds of light must 

 show themselves for which at the respective temperature the 

 power of emission is above zero. But since the incandescent gas 

 is always of relatively small density, all the differences in the 

 emission power of the various kinds of light must become 

 apparent in the spectrum, and thus the latter must be richly 

 varied or shaded ; this is indeed the case in the band-spectra of 

 gases. Prof. Wiillner adduces the spectra of iodine vapour as 

 proofs of the correctness of his view. When rendered incan- 

 descent by means of a hydrogen flame, iodine vapour gives the 

 negative absorption spectrum, which is of the same character as 

 the band-spectra of gases ; if rendered incandescent through the 

 spark, the glowing iodine molecules give a bright line-spectrum. 



The band-spectrum of nitrogen shows, that this element at 

 the temperatures obtained by electric discharges pos-esses quite as 

 great a power of absorption as that of iodine vapour at low- 

 temperatures, because the band-spectrum of nitrogen is essen- 

 tially of the same character as that of iodine vapcj ur, however 

 different it may be from it in detail. Of all other gases, nitrogen 

 must therefore be particularly adapted for showing, through "he 

 examination of the light it emits, the dependence of spectral 

 phenomena from the thickness and density of the radiating layer 

 of gas, and thus for furnishing the proof that there is no constant 

 spectrum of nitrogen, but that a certain spectrum exists only at 

 a certain temperature and density of the gas. This indeed is 

 the question upon which turns the difference of opinions of Ang- 

 strom and Lockyer on the one hand and of Wiillner and Zollner 

 on the other ; the former ascribing the different spectra to 

 chemical differences in the molecular conditions of the gas, the 

 latter merely to differences of temperature, density, and thickness 

 of the radiating layer. 



In a former treatise on the nitrogen spectrum. Prof. Wiillner, 

 without having recognised the importance of the density of the 

 radiating layer with regard to the light emitted, pointed out that 

 when the pressure in a nitrogen tube is diminished to such an 

 extent that it ceases to be measurable, the brightness of the 

 spectrum decreases, and in such a manner that the darker parts 

 first fade away, so that at last only the brightest parts remain. 

 He added that in this way the nitrogen spectrum in its character 

 approaches a spectrum of the second order (the name given to 

 line spectra by Pliicker) without, however, changing to the nitro- 

 gen spectrum of the second order, since no new bright lir.es 

 appear. At that time, however. Prof. Wiillner did not continue 

 his researches in that direction, and in particular he did not 

 examine whether the bright parts remaining do indeed correspond 

 to the maxima of the complete band spectrum, because the 

 spectra he obtained at those pressures were too weak to allow of 

 measurements being made with the instruments then at his dis- 

 posal. Lately, however, the Professor has minutely examined 

 the nitrogen spectrum in this stnse, employing a simple contriv- 

 ance for rendering the spectra bright enough for measurements, 

 even at the lowest pressures. This consisted in the employment 

 of spectral tubes of very narrow calibre (about 2 mm. in 

 diameter). 



It must be remembered here that the temperature of the gas, 

 which is caused by the induction current, rises with decreasing 

 diameter of the tube. (If, however, the tubes were taken tio 

 narrow, the current at once-broke them. ) 



Since the resistance in the tube rises as the density of the gas 

 decreases, at least from a certain point of low pressure down- 

 wards, the temperature rises as well. If the rise in the tempera- 

 ture was sufficiently great, the experiment described by Prof. 

 Wiillner necessarily decided the only hypothetical part in his 

 conception of spectral phenomena, viz., whether with a rising 

 temperature the absorption power for the various kinds of light 

 grows in a similar manner or not. If it does grow simultaneously 

 with the temperature, then the relative maxima of intensity of 

 light which the complete band spectrum shows must always re- 

 main the same ; the bright parts remaining at the lowest densi ty 

 must correspond to the maxima of the band-spectrum. If the 

 contrary is the case, and this is what Prof. Wiillner assumes, 

 then dark parts in the band-spectrum may become the brighter 

 ones as the density decreases, and the bright parts remaining at 

 the loivest pressure may be situated at places in the spectrum 

 differing widely from the maxima of the band-spectrum. The 

 first part of the experiment therefore consisted in an exact deter- 

 mination of the relative maxima in the band-spectrum of nitro- 

 gen for the sake of comparison. This is minutely described in 

 Prof. Wiillner's paper. The final results of the observations 

 were in complete accordance with the Professor's conception of 

 the spectral phenomena. There is indeed no definite nitrogen 

 spectrum when in layers of sufficient thinness the density of the 

 gas is reduced below a certain limit. The band spectrum changes 

 step by step into a line-spectrum ; this, however, is not identical 

 with the line-spectrum produced by the spark, but has only a 

 certain number of lines in common with it. In this gradual 

 change it is easy to follow the disjilacement of the maxima of 

 brightness which takes place little by little as the temperature 

 rises, and is quite conspicuous in several places ; this displace- 

 ment is the very cause why in this line-spectrum the lines in 

 places differ widely in their situation from the maxima of 

 brightness of the baud spectrum. 



Prof. Wiillner then gives an exact description of that part of 



