On the Relative Intensity of the Spectral Lines of Gases. 329 



more or less. Kefilling the water-flasks had also a disturbing 

 effect. Taking these facts into consideration, it cannot be 

 wondered at that at this temperature the results were rather 

 inconstant. The tubes gave, when filled with n NaCI solution, 

 at 15° an amount of salt represented by 1*1 cubic centim. 



40 AgN0 3 per hour. But usually, when in the ice-chest, the 



salt diffused from similar solutions required for its titration 



1*5-2 cubic centims. ^AgN0 3 per hour, indicating that the 



progress of the diffusion was disturbed. On one occasion, 

 however, for a period of about 14 hours, 6 P. M. to 8 a.m., the 

 diffused amount was equivalent to 0*87 cubic centim. per hour. 

 This rate did not continue during the next day ; opening the 

 box to renew the ice and water disturbed it as usual. The 

 experiments at a higher temperature were somewhat more 

 successful. Instead of the ice a Bunsen burner was placed in 

 the box, and a tolerably constant temperature of 25° was ob- 

 tained. The amount of AgN0 3 solution required now per 

 hour varied between 1*3 and 1*4 cubic centim. Assuming 

 that for the lower temperature the result 0*87 cubic centim. 

 was correct, the amounts diffused from solutions of the same 

 strength at three different temperatures may be thus tabu- 

 lated : — 



At 10°, 0*87 cub. cent. ~ AgN0 3 required per hour. 



15° 1-10 



33 xo 3 x ±KJ ?3 33 13 ?? 



„ 25 , 1*35 „ „ „ „ 



But these results can only be considered approximately 

 correct. 



[To be continued.] 



XLII. Relative Intensity of the Spectral Lines of Gases. 

 By J. Rand Capkon, F.R.A.S* 



IN reference to M. Ch. Fievez's " Researches on the Rela- 

 tive Intensity of the Spectral Lines of Hydrogen and 

 Nitrogen in regard to the Constitution of Nebulae," which 

 appears in your present month's Number, and a copy of which 

 the author has obligingly sent me, it may perhaps be useful 

 to refer to a few experiments detailed in my 'Auroras and 

 their Spectra,' chapter xi. p. 108 et seg. 



These were made for the purpose of testing how far it was 

 possible or probable that the bright green and red lines in the 

 aurora-spectrum might be the brighter and only visible rays 

 * Communicated by the Author. 

 Phil. Mag. S. 5. Vol. 9. No. 57. May 1880, 2 B 



