GASEOUS SPECTRA UNDER HIGH DISPERSION. 447 



rarefied, medium; but evidently the same linear spectrum in each case, and of 

 but a few, say a score or two, lines even at brightest.* 



With the simple or direct spark on the contrary in the open air, those two 

 spectra vanish, and are replaced by other two, perfectly different; whereof, as 

 just described, that of O is barely visible to moderate power, even in its isolated 

 compound triples or the stronger of its single lines lately discovered ; while N 

 is heavily conspicuous all along the spectrum in the shape of a closely packed 

 arrangement of numerous, narrow bands. In a vacuum tube of N alone this 

 arrangement is still more brilliant, is generally known as " the band spectrum 

 of N," and " is one of the most beautiful," says the British Association Report, 

 " which can be observed." 



The best map of it I have yet seen is that in Angstrom and Thalen's paper 

 of 1875; a map about 27 inches long, and exquisitely engraved; i.e., so far as 

 engraver's work alone is concerned, for the vertical lines wherewith the bands 

 are shaded are engraver's ornament only, and have no pretension to represent- 

 ing lines seen by the observer. But with this reservation accepted, the map 

 effectively reproduces all that was known of the spectrum until 1880; when 

 Prof. Alex. S. Herschel communicated to this Society some notable exten- 

 sions of the spectrum into the ultra red, which he had just then made with my 

 Spectroscope as it then was, or merely in a pretty good condition. 



His whole conclusion I believe was, that he could identify many more of 

 the narrow bands of N, in that lower region, far beyond or outside the place 

 where the first of Thalen's pictured bands begin. And he could even trace 

 them up to a place where a triplet of sharp lines shot up, and seemed to form 

 a sort of fountain head, whence had flowed down the continued stream of 

 Nitrogen cross-bands all through the Red, the Orange, the Yellow, and the 

 Green of that spectrum. 



With my present improved instrument, and several new and exquisitely pure 

 tubes of N prepared for me by both M. Demichel in Paris, and Mr Casella in 

 London, at various pressures between TO" and 25", — I have been enabled 

 somewhat to modify the above view, as thus, — 



(1) The band spectrum of Nitrogen, even in those brilliant colour regions 

 just cited, is by no means one uninterrupted series of similar bands, — but is a 

 succession of four large groups of bands; each group bringing in with it slight 

 variations on the preceding one, and separated from its neighbour group on 

 either side, by a tolerably distinct breadth of 2 or 3 bands of weaker action. 



(2) Thalen's first band, though no longer to be regarded as the first band 

 or beginning of the N spectrum, is yet the first band of its own, or the Red, 

 group; which we may therefore worthily denominate Thalen's group. 



* The manner in which the Red Hydrogen line comes into that Spectrum, is very striking, and I 

 have not yet seen a good reason given for it. 



VOL. XXXTI. PART TIL 4 C 



