318 Sir John Conroy on the Change in the 



p. 174) that he " was surprised to observe that it produced 

 opposite effects upon different glasses, diminishing the ab- 

 sorbing power in some, and increasing it in others." He 

 found that the transparency of a piece of purple glass was 

 much increased on heating, whilst that of a yellowish-green 

 glass and of a red glass was diminished ; the purple glass 

 recovered its colour on cooling- the other two did not do so 

 completely. 



Feussner (Fort schritte der Physik, 1867, p. 237) made 

 some observations on the effect of heat on the absorption- 

 spectra of substances in solution. 



No observations on the effect of heat on the transparency 

 of solid substances for rays of different refrangibilities except 

 those of Sir David Brewster appear to have been published, 

 although, of course, the change of colour which borax blow- 

 pipe-beads containing certain metallic oxides undergo on 

 cooling, is well known. I therefore venture to communicate 

 to the Physical Society some determinations which 1 have 

 recently made on the changes produced by heat in the 

 absorption-spectrum of cobalt glass. 



The absorption- spectra of coloured glasses are not, as a rule, 

 very characteristic, and merely show a continuous absorption 

 extending through a considerable portion of the spectrum. 

 Cobalt glass, however, as is well-known, has a characteristic 

 absorption-spectrum consisting of three dark bands, in the 

 red, yellow, and green, with a considerable amount of absorp- 

 tion between the first two ; so that with a rather deeply 

 coloured glass the transmitted light consists merely of the 

 extreme red, some yellowish-green, and the blue and violet 

 rays. 



A small piece of this glass was heated by means of a 

 medium-sized Bunsen (15 millim. tube), the glass being 

 supported on, and nearly surrounded by, combustion-furnace 

 tiles. It was found that in this way the glass could be heated 

 till the edges began to soften and were visibly red, without 

 much risk of its cracking either whilst being heated or during 

 its subsequent cooling. An ordinary gas-jet was used as the 

 source of light, and the light transmitted by the glass 

 examined with a spectroscope ; a small direct-vision one 

 being used, as it was found that the changes in the spec- 

 trum were less distinctly seen with a spectroscope of greater 

 dispersive power. 



The absorption-spectrum of the cold glass is represented in 

 fig. 1, which is drawn to an arbitrary scale, and not to one of 

 wave-lengths. On gradually heating the glass, the absorption 

 between the two first dark bands, those in the red and yellow, 



