156 Intelligence and Miscellaneous Articles, 



posal permit ; and I hope to be able soon to make known the entire 

 spectra of the principal simple gases. — Comptes Rendus de VAca- 

 demie des Sciences, May 21, 1877, tome Ixxxiv. pp. 1151-1154. 



PHOTOMETRIC RESEARCHES ON COLOURED FLAMES. 

 BY M. GOUY. 



To continue these researches, I had, in the absence of any data 

 on the subject, to make a preliminary study of the conditions on 

 which the brightness of coloured flames depends : the most impor- 

 tant are the thickness of the flame, its composition, the nature of 

 the salt, and the quantity of it carried along in the combustible 

 mixture. The apparatus employed has been described in the 

 Comptes Mendus*. 



1. The augmentation of brightness of a line when the thickness 

 of the flame becomes twice as great is easily deduced from the ex- 

 periments described in a previous Note. That result, controlled 

 and completed by another method, leads to this relation : — AVhen 



the thickness of the flame is augmented by a fraction -, the bright- 



Ic 

 ness of the line increases by the fraction - ; ^- is equal to 0*35 for 



sodium, 0-45 for lithium, and is comprised between 0-9 and unity 

 for the bands of calcium and strontium. This coefficient diminishes 

 a little in proportion as the brightness of the flame increases. The 

 formula is inapplicable for ?i<:l ; it supposes also that the flame is 

 homogeneous — which can be secured by viewing it a little above the 

 orifice. 



2. If the flame is a little overcharged wdth illuminating gas, and 

 this be gradually diminished, the brightness is seen to increase, 

 pass through a maximum, and rapidly diminish. With lithium the 

 augmentation of brightness is scarcely noticeable, the maximum 

 takes place before the flame ceases to be reducing (for a copper 

 w^Li'e), and is follow^ed by a rapid diminution. AVith a large excess 

 of air the spectrum disappears. Calcium, barium, and strontium 

 behave in the same manner ; this fact appears unfavourable to the 

 opinion which attributes the bands of their spectra to oxides. 

 With sodium, on the contrary, the brightness of the flame augments 

 rapidly as it becomes less reducing, the maximum is produced at 

 the iustant w^hen it ceases to be so, and is followed by a reduction 

 of brightness much slower than with the other metals — so that a 

 flame containing soda and lithia is red with an excess of illumina- 

 ting gas, and yellow with an excess of air. 



These results are confirmed by other experiments made with 

 one and the same flame, which can at will be made to burn in air 



* See t. Ixxxiii. p. 269, and t. Ixxxiv. p. 231 (Phil. Mag. [5] vol. ii. 

 p. 317, & vol. iii. p. 238). 



