Analysis of the Spectrum of Sodium. 503 



To plot out the curve giving the mixtures of two spectrum 

 colours, take a third spectrum colour and make a colour 

 patch of the first two colours, and another colour patch of the 

 third colour and white. Keep the luminosity of the third 

 colour constant, and vary that of the other colours and the 

 white until both patches are of the same colour. Then measure 

 the quantity of white used and mark off a line from the 

 position of the third spectrum colour to the right propor- 

 tional to the quantity of white. The end of this line gives a 

 point in the curve. By taking other spectrum colours as the 

 third colour other points may he obtained. If, however, it is 

 found impossible to make the two patches of the same colour, 

 then throw the three spectrum colours together, and keeping 

 the luminosity of the third colour constant vary that of the 

 other two until the three produce white ; measure the quantity 

 of white, and mark off to the left from the position of the 

 third spectrum colour a line proportional to the quantity of 

 white obtained. The end of this line is a point in the curve. 

 If both these methods fail the point on the curve correspond- 

 ing to the third spectrum colour lies in the abnormal or 

 imaginary regions, and cannot be determined by experiment. 



When the derived curves have been plotted out by experi- 

 ment, it will be possible to modify the hypothetical forms of 

 the curves of intensity of the sensation in the spectrum so as 

 to make the curves derived from them accord more closely 

 with the results of experiment, and so to arrive by gradual 

 approximation to the true form of those curves. 



LVIII. Analysis of the Spectrum of Sodium, including an 

 Inquiry into the true place of the Lines that have been 

 regarded as Satellites. By Gr. Johnstone Stoney, M.A., 

 D.Sc, F.R.S., Vice-President, Royal Dublin Society*. 



[Plates VI. & VII.] 



Introduction. 



N analysing spectra, the following definitions and distinc- 

 tions will be found of use : — 



j, the jot of time, is the time that light takes to advance 

 one tenth of a millimetre in the open aether. In the 

 open aether all rays f travel this distance in the same time, 

 which is approximately one third of the twelfthet (one 



* Communicated by the Author. From the Scientific Proceedings of 

 the Koyal Dublin Society, vol. vii. part 3. 



t By a ray of light is meant the light producing a line in the spectrum. 



I 



