60 Messrs. Paterson and Dudding on Estimation of 



the black body and grey body at the same apparent tempe- 

 rature (measured optically) . It is this difference of relative 

 distribution which results, by the colour-identity method, in 

 the grey body (curve G) being given a temperature of 

 1750° C. and the black bodv (curve B) a temperature of 

 1578° G. 



Now, the curve for platinum at its melting-point lies close 

 to curve G for a grey body. Waidner and Burgess's platinum 

 melting-point determinations for the three wave-lengths 

 given in Table VI. are plotted in curve D. If the figures 

 published by Waidner and Burgess may be depended on to 

 give the relative intensities in the three wave-lengths, curve D 

 indicates that the colour-identity method should have given 

 a value for the melting-point of platinum above 1750° G., 

 since, as compared with the grey body, Waidner and Burgess 

 show relatively more radiation from platinum at the blue 

 than at the red end of the spectrum. Before definite con- 

 clusions can be drawn it would be desirable to have measure- 

 ments in other than the three wave-lengths considered and 

 information as to the monochromatism of the light in each 

 of the wave-lengths for which the intensities are plotted in 

 curve D. The figures for the melting-point of platinum , 

 shown in Table V., indicate, it is true, a tendency to fall in 

 the direction to be expected from Waidner and Burgess's 

 values plotted in the diagram, but not, however, as much as 

 Line D indicates. The opinion, therefore, expressed by Hyde 

 that the colour-identity method will err in ascribing tempe- 

 ratures which are, if anything, slightly too high, is supported 

 so long as the bodies in question are selective in the visible 

 spectrum in favour of the shorter wave-lengths. If they are 

 selective in favour of the red end, the temperature ascribed 

 will tend to be low, whilst if they are true grey bodies the 

 method will be accurate. 



It should, however, be recalled that compared with other 

 substances platinum is regarded as a relatively selective 

 body in the visible spectrum, and if this is so, the differences 

 indicated in fig. 6 are for a fairly extreme case. 



Little is actually known as yet regarding the depariure of 

 metallic bodies from the characteristics of grey bodies in the 

 region of the visible spectrum. Any deviation which there is 

 would seem to be small in amount and insufficient to invali- 

 date estimations of the temperature by the colour-identity 

 method, intelligently used. It is suggested that the method 

 should be specially useful for assisting in the determination 

 of temperatures and melting-points of some of the more 

 refractory substances whose true temperatures by the usual 



