and of Simple Substances. 187 



point on the barium spectrum, that I could observe a distinct 

 line in the spectrum of this metal. This line only becomes 

 visible by methods 6 and 7 (electricity), or by bringing barium 

 compounds into the flame of cyanogen burning in oxygen. 



Such a relation, as in the case of calcium, strontium, and 

 barium compounds, could not be discovered in the other metals. 

 Similarity in the spectra of the compounds of a metal is easily 

 seen in the case of copper — for instance, in the violet part of the 

 bromiile- and chloride-of-copper spectrum, and in the green part 

 of the chloride-, bromide-, and iodide-of-copper spectrum ; but no 

 relation could be established between these spectra and the other 

 properties of the compounds. 



As there are resemblances in the spectra between individual 

 compounds of a metal, so there are such between compounds of 

 different metals with oxygen. Most striking is the resemblance 

 between the spectra of lime and of strontia ; the individual parts 

 correspond to each other, but the lime spectrum is more extended 

 than that of strontia. 



The spectra of baryta and oxide of lead are also very similar 

 in certain respects. This similarity is difficult to express in the 

 drawings, while it is very striking to the sight. 



Relations between these spectra and the atomic weights I have 

 not as yet been able to establish, but it may be expected that 

 such exist. 



I have not been able to find properties of the spectra of the 

 metals which enable a connexion of the metals with one another 

 to be recognized : similarity of individual spectra together with 

 similar properties of the metals seems to point at such a con- 

 nexion; thus, for instance, the spectrum of zinc is very like 

 that of cadmium. 



The metalloids show the same spectra, provided with regular 

 shading, as the metallic oxides ; and if, for instance, the spectrum 

 of the oxygen compounds of bismuth is compared with the spec- 

 trum of iodine obtained by method 6 (volatilization in the glass 

 tube), a resemblance is observed corresponding to that which I 

 have noticed as existing between the spectrum of oxide of lead 

 and of baryta. The oxides of both these metals are decomposed 

 by high temperatures, and show individual lines as spectra. 

 Iodine also, when investigated by method 7 at a high tempera- 

 ture, exhibits an entirely different spectrum, consisting of indivi- 

 dual lines like the spectra produced at high temperatures from 

 these metallic oxides; and from the phenomenon that iodine 

 shows two spectra, from their similarity with the spectra of the 

 metallic oxides and those of the metals, and from its ana- 



02 



