154 Intelligence and Miscellaneous A^'ticles, 



At the ordinary pressure, in nitrogen, the spark is pale and 

 ridged with little strokes of fire; in the spectroscope, upon the 

 wires are seen the channellings attributed to nitrogen, and in the 

 intervals between these the principal lines of that gas. On com- 

 pressing, one sees the channellings gradually fade, while the lines 

 grow fainter and the continuous ground of the spectrum becomes 

 brighter. At t^'o atmospheres there are only six nitrogen-lines 

 from the orange to the blue, and five diffuse bands beyond. At ten 

 atmospheres there remain only two nitrogen -lines, \=567 and 

 \'=:500, and then a brilliant line in the violet (/\"=424), which 

 made its appearance at five atmospheres, and which from my last 

 experiments I attribute to nitrogen. The sodium-line is very dis- 

 tinct, while it was not to be distinguished at the ordinary pressure, 

 which makes evident the part played by the wall of the tube. 

 About 15 atmospheres the spark becomes dazzling ; upon the con- 

 tinuous spectrum the four preceding lines are caught sight of, and 

 some bright points which are due to platinum. The pressure was 

 carried up to 40 atmospheres without these peculiarities ceasiug to 

 be distinguished. 



When the gas was reduced to the ordinary pressure, the spectrum 

 resumed its primitive aspect, but the sodium-line persisted upon 

 the negative electrode ; when the current was reversed, this line 

 instantaneously passed from one electrode to the other, as if a so- 

 dium-compound had been electrolyzed. Purther, in air compressed 

 above 30 atmospheres the spark produced intense shining vapours, 

 and the spectroscope showed the absorption spectrum of hyponi- 

 trous acid. It was a very fine spectrum, although the absorbing 

 layer was less than 3 miilims. in thickness. 



The observation of the above phenomena is very fatiguing ; and 

 it is impossible to seize all the details, to reproduce them in an ac- 

 curate drawing. The irradiation-effects dim the sight ; and judg- 

 ments upon the real condition of the spectrum cannot be correct. 

 Photography permitted these inconveniences to be avoided. To it 

 I had recourse, and succeeded in obtaining plates sufficiently deli- 

 cate to favour precise measurements ; and at the same time I found 

 my opinion, on the state of the electric spark in a gas, confirmed 

 in a striking manner. 



I made use of an ordinary spectroscope with only one flint-glass 

 prism, replacing the eyepiece by a small dark chamber. I operated 

 upon either wet or dry collodion, according to the duration of the 

 pose, which A^aried from fifteen minutes to one hour. The spectrum 

 is photographed from the line F to the line M ; and its intensity 

 between the lines Gr and L is remarkable. I recognized the prin- 

 cipal lines by photographing side by side on the same plate the 



solar spectrum and that of the spark, taking advantage of Angstrom's 

 Plates completed by Mascart and Cornu for the former. 



The following, for example, are the results relative to the part of 

 the nitrogen-spectrum comprised between G and H. The gas is 

 at the ordinary pressure, in a metal cylinder with a glass plate ; 



