448 On the Spectra of Chemical Compounds. 



A glass tube 80 millims. in length and 15 millims. in width 

 was filled with bromine vapour ; no liquid bromine was pre- 

 sent at the ordinary temperature. The lines from — 1 near 

 D to +23 in the greenish-blue were observable. The tube 

 was warmed by means of a Bunsen's lamp : the lines —2, —3, 

 —4, to —6 became visible; the area of total absorption ad- 

 vanced from the blue end until it reached the line +12. The 

 temperature of the bromine was determined by observing the 

 melting of little pieces of various metals brought into the 

 tube, to be about 350°. Increasing the temperature had thus 

 caused the lines from —2 to —-6 to become visible. I have 

 already stated that increasing the thickness of the bromine 

 layer also caused lines in the red to appear : I found by actual 

 experiment that an increase in temperature from 20° to 350° 

 was equal, in this respect, to increasing the thickness of the 

 layer from 15 to 16 millims. 



Analogous results were obtained with the spectra of iodine 

 and of nitrogen tetroxide. 



The following observations regarding effects of increased 

 temperature relate to individual lines in each of the three 

 spectra examined : these effects were not produced by increa- 

 sing the thickness of the absorbing layer, but only by increasing 

 the temperature. The two lines referred to diminished in in- 

 tensity, and finally disappeared, when the temperature was in- 

 creased, while the intensity of the neighbouring lines, as also 

 that of the whole spectrum, was increased. 



Bromine, iodine, and nitrogen tetroxide were warmed in 

 sealed tubes 80 millims. in length and 15 millims. in width : 

 the following lines disappeared: — 



(1) In the bromine-spectrum. 



(a) The line +12 ; the space between 11 and 13 gradu- 

 ally became filled with fine lines. 



(h) The line + 15 ; the space between 14 and 16 became 

 filled with fine lines. 



(2) In the iodine-spectrum. 



A line nearly in the same position as the bromine-line 15, 

 the space between the two neighbouring lines becoming 

 filled with fine lines. 



(3) In the nitrogen tetroxide spectrum. 



Three lines close by C, standing out some distance from 

 one another. 

 Weiss* says that he noticed a change in the position of cer- 

 tain lines in the absorption-spectrum of nitrogen tetroxide, pro- 

 duced by increasing the thickness of the layer of gas, or by rais- 

 ing the temperature. My own experiments have convinced me 

 * Pogg. Ann. cxii. p. 155. 



