462 



From the similarity of these formulag with Dulong and Petit's law of heat- 

 radiation, Becquerel regards them as being confirmed by analogy. The de- 

 terminations of the temperatures in his experiments were all deduced from the 

 intensity of the thermoelectric current of a platinum-palladium junction, 

 and the luminous intensities were determined by means of a photometer based 

 on double refraction. 



The observations made in connexion with this Report on the increase of total 

 luminous intensity have been conducted similarly to those detailed by Draper in 

 the Philosophical Magazine for 1847. The apparatus has been modified so as 

 to be more conveniently employed, and the experiments made were found on 

 being tabulated to be very well expressed by the following empirical formula : — 



990 + 71 46 =n= I, 



where I is the luminous intensity, and 990-|-n46° is equal to the total tem- 

 perature — that is to say, above the temperature of 1036° C, by which time 

 all the luminous rays may be considered present ; the intensity is a parabolic 

 function of the temperature. The curve of increase is therefore a very acute 

 parabola. The diagram, p. 463, contains the curves of Becquerel, both for 

 homogeneous rays and for white light, and also the curve given by the above 

 formula. It is evident the rate of growth of the total luminous intensity is 

 very much slower than that obtained by Becquerel. The curve resembles the 

 rate of growth obtained from the homogeneous rays in his observations, although 

 all his curves begin more slowly and finish with far greater rapidity. This 

 doubtless depends on the thermometric degrees diminishing rapidly with 

 higher temperatures, according to his plan of measurement ; but the great 

 variation in the curves when taken, even for the same kind of ray, shows that 

 little reliance can be placed on the results. 



As the observations on increase of luminosity above 1000° C. can only be 

 carried on for a range of 500° C. with the expansion of platinum, it was very 

 essential that some comparison between the results of the empirical law given 

 above and actual observation should be made at higher temperatures. For 

 this purpose, a series of observations were made as to the relative light-inten- 

 sity of lime heated to a temperature of 2000° C. in the oxyhydrogen flame, 

 and the same substance at the boiling-point of zinc, temperature 1040° C. 

 The following plan was adopted in making observations : — A square pencil of 

 lime, four or five millimetres on the side, and of a length of 50 miUims., was 

 supported horizontally, and the inner cone of a powerful oxyhydrogen flame 

 was made to play on a smooth cross section of the pencil. The light emitted 

 from this perpendicular surface had to pass through a small circular aperture 

 into an adjoining dark chamber for the purpose of comparison with the light 

 emitted from an equal surface of lime, the temperature of which was near the 

 boiling-point of zinc. In order to get a temperature maintained near 1000° C, 

 I have adopted the following method : — A piece of platinum of an equal surface 



