ON OUR KNOWLEDGE OF SPECTRUM ANALYSIS. 141 



The complete spectrum of iodine is placed most towards the blue, 

 that of chlorine is placed most towards the red end of the spectrum. 



In a subsequent paper Ciamician deals further, and we believe more 

 successfully, with the question. He begins by examiuiug the spectra of 

 the carbon compounds. The spectrum of cjanogen consists of two sets 

 of bands, one in the red and one in the blue ; the bands in the red 

 resemble the nitrogen bands, the bands in the blue resemble the carbon 

 bands (candle spectrum). Ciamician therefore suggests that the bands 

 in the red are due to the nitrogen atoms in the cyanogen molecules, while 

 the bands in the blue are due to the carbon within the same molecule. 

 There can be no doubt as to the correctness of the resemblance pointed 

 out by Ciamician, and the only argument which can be urged against bis 

 conclusion is the uncertainty that both sets of bands belong to the same 

 compound of carbon and nitrogen. There are some reasons for supposing 

 that they are not ; but as yet this is an open question. The spectrum of 

 carbonic oxide is very similar to that of the carbon (candle spectrum), as 

 has often been shown. 



Ciamician makes a bold use of the division of the cyanogen spectrum 

 into two halves, which, as just pointed out, resemble the spectra of nitro- 

 gen and carbon respectively. For he thinks that we can divide the 

 spectra of some of the elements similarly into two parts, one part re- 

 sembling the spectrum of one, the other part resembling the spectrum of 

 a second element, and this division he suggests may be due to the fact 

 that such an element is really a compound body, wlaich is composed of 

 the two elements to whose spectrum its own is analogous. The high 

 temperature spectrum of silicium consists of a series of lines. The most 

 refrangible half of the lines certainly resembles in its general arrange- 

 ment to a marked degree the lines of carbon. The less refrangible lines 

 Ciamician believes to resemble some of the oxygen lines, but this analogy 

 does not appear very strikingly in the map which accompanies his paper. 

 Silicium at low temperatures shows bands which, as pointed out by 

 Messrs. Troost and Hautefeuille are very much like the carbon bands. 

 Ciamician has also for the first time succeeded in obtaining two 

 spectra of boron. Both the band and the line spectrum show a strikino* 

 resemblance to the carbon spectra. The line spectrum of aluminium 

 resembles in its most refrangible half also that of carbon, while its less 

 refrangible half resembles the less refrangible half of silicium. The com- 

 plete similarity of the spectra of aluminium and silicium, and that likeness 

 of their most refrangible parts with the complete spectra of carbon and 

 boron is the most striking analogy which has as yet been pointed out ; 

 but it ought perhaps to be added that Ciamician's drawing of the boron 

 spectrum does not altogether tally with the account given by Messrs. 

 Troost and Hautefeuille. The line spectrum of magnesium also shows a 

 likeness to the line spectrum of carbon ; and the carbon bands (candle 

 spectrum) resemble certain bands seen in the spectrum of magnesium 

 under certain circumstances, which according to Liveing and Dewar 

 always involve the presence of hydrogen.' 



' An example may here be given as to the misunderstandings which may arise 

 owing to the confusion which reigns at present in spectroscopic nomenclature. 

 Messrs. Dewar and Liveing have pointed out the resemblance of the hydrocarbon 

 spectrum and the hydrogen-magnesium spectrum (Proc. Boy. Soc. xxx. p. 161, 1880). 

 Ciamician writes that he also discovered the resemblance ; but what Ciamician calls 

 the hydrocarbon spectrum is an altogether different spectrum, presenting no resem- 



