248 Professor George D. Liveing [Marct 9, 



with compounds. Nitrogen gives in the arc as well as in the spark 

 a channelled spectrum of singular beauty, extending with but short 

 breaks almost to the extremity of tlie ultra-violet region which we have 

 examined. (See photograph 7.) These multitudinous lines of nitrogen 

 constantly present in the arc taken in air, help to make the problem 

 of unravelling the spectrum of the arc, and assigning each line to its 

 proper source, far more difficult than it might at first sight be supposed. 

 Carbon, which in the arc frequently gives a channelled spectrum in 

 the visible region, gives only a limited number of lines in the ultra- 

 violet ; but cyanogen gives one set of flutings near the line L, and 

 another near N, which are so brilliant in the arc as to obscure the 

 metallic lines in their neighbourhood (photograph 4). To the same 

 class we may refer the spectrum of water, of which the most brilliant 

 portion is given in photograph 6. 



The series of lines produced by the same element, which I have 

 spoken of as overtones of a fundamental vibration, have been likened 

 to these channellings, but in reality they are very different. In the 

 series, which I have supposed to have a sort of harmonic relation, the 

 successive lines or groups of lines invariably become nearer to one 

 another as the wave-lengths become shorter, and at the same time 

 they diminish in strength and sharpness ; whereas in the channelled 

 spectra the strongest lines are at the end where they are most closely 

 set, and they generally diminish in strength as they get further apart. 

 Also increase of distance between the lines of channelled spectra is 

 sometimes towards the less, sometimes towards the more, refrangible 

 end of the spectrum. 



I have before observed that a great part of the ultra-violet spectra 

 of the elements which we have observed lies entirely beyond the limit 

 of the solar spectrum ; that limit is the line U at wave-length 2947. 

 But though this is the limit of the solar radiation which reaches us 

 on the earth, we can hardly suppose that the sun itself, or the photo- 

 sphere, emits no radiation of shorter wave-length. We know that 

 there is plenty of iron and magnesium in the sun, and the strongest 

 radiations at high temperatures of these elements are of shorter 

 wave-length than U. Moreover, the continuous spectra of incan- 

 descent solids in many cases extend far beyond U. The continuous 

 spectrum of burning magnesium reaches quite up to the wave-length 

 2380, that of the flame of carbon disulphide mixed with hydrogen 

 and fed with oxygen reaches even further, that of lime heated with 

 an oxyhydrogen blowpipe, though feeble beyond the limit of the 

 solar spectrum, extends up to wave-length 2680. The tempera- 

 ture of the sun cannot be less than that of any of these sources 

 of heat, so that we are forced to suppose that the radiation, more 

 refrangible than U, which leaves the body of the sun, is stopped 

 somewhere either in our atmosphere, or in planetary space, or in 

 the atmosphere of the sun himself. Now Cornu has found that 

 when the thickness of our atmosphere traversed by the sun's rays 

 is diminished as much as possible by taking the sun at its greatest 



