402 Messrs. Nichols and Snow on tlxe Influence of 



sorption of light with elevation of temperature, the more 

 refrangible increment increasing at a greater rate than the 

 less refrangible." How accurate a statement of fact this is 

 will appear when the measurements to be given in the present 

 paper have been described. 



The kindred topic of the influence of temperature upon 

 absorption-spectra has claimed the attention of Gladstone*, 

 Bartley f, and, very recently, of Conroy J. Hitherto, however, 

 there have been but few efforts to analyse systematically 

 the selective reflexion which takes place at the surface of 

 pigments §, and no attempts to trace out quantitatively those 

 variations in the amount and character of the reflected light 

 which are brought about by the action of heat. 



It is our purpose in this paper to present the results of an 

 investigation of the colour-changes which pigments undergo 

 when subjected to wide ranges of temperature. The deter- 

 minations included the spectroscopic analysis of the light from 

 the pigment, with means of measuring its intensity, wave- 

 length by wave-length, and the estimation of the temperature 

 of the coloured surface. 



The method of determining temperature which seemed to 

 present the fewest objections, consisted in supporting a thin 

 layer of the pigment on a ribbon of platinum foil, which could 

 be maintained at any desired temperature by means of the 

 electric current, and then measuring the linear expansion of 

 the foil. To this end a strip, 25 centim. in length and 

 7 millim. wide, was cut from a sheet of platinum. The strip 

 was placed in the circuit of a Gramme dynamo, and heated 

 to a cherry-red. The degree of incandescence, which was 

 observed to be in all parts the same, afforded a sufficient test 

 of the uniformity of the foil. To the ends of this platinum 

 strip were soldered brass rods, capable of carrying a heavy 

 current without heating. These passed through the arms of 

 a frame in which the strip was mounted, and one of them bore 

 a compressed spiral spring, which served to keep the platinum 

 ribbon properly stretched. By means of a dividing-engine 

 two fine diamond rulings, 89 millim. apart, were drawn upon 

 the foil at right angles to its length. Two microscopes, each 



* Philosophical Magazine, [4] xiv. p. 423 (1857 > 



t Proceedings of the Royal Society, xxii. p. 241. 



% Philosophical Magazine, [5] xxxi. p. 317 (1891). 



§ With the exception of a few measurements made several years ago 

 by one of us (Nichols, American Journal of Science, vol. xxviii. p. 343), 

 the only quantitative study of the spectra of pigments with which we are 

 acquainted is that contained in the recent paper of Abney and Festing on 

 Colour Photometry (Phil. Trans, clxxix. p. 549). 



