Pluok&r on Spectrum Analysis. 267 



spectrum ; on the other hand, it results that if we employ con- 

 siderable magnifications, and augment the refraction, the bands 

 in question will remain distinctly portrayed, while the continuous 

 spectrum and the coloured spaces become almost imperceptible. 

 Thus, when employing the telescope of the great spectroscope 

 apparatus of Steinheil, I immediately perceived that.. the bands 

 of the homogeneous light which I admitted into the violet part 

 of the nitrogen spectrum only existed under the cbnditions 

 cited, while, in the case of hydrogen, I was able to confirm the 

 existence, in the obscure part of the spectrum, of homogeneous 

 bands of very feeble intensity. 



" When, for the sake of giving a more elevated temperature 

 to a rarefied gas, I caused the current occupying a larger space 

 to pass through the capillary tube, I have observed, from the 

 commencement of my researches, a change of colour accom- 

 panying the change of intensity. In other words, the relative 

 luminous intensity of the different homogeneous lines which 

 usually constitute the spectra of gases, is seen to be a function of 

 temperature. I subsequently showed that in the case of hydro- 

 gen, the intensity of the three lines forming the essential part 

 of its spectrum, do not diminish in equal proportion, and that 

 the red line is extinguished first as we approach through the 

 rarefaction of the gas to the point beyond which it cannot 

 transmit the electric current. Latterly many experiments ap- 

 peared to contradict my former observations, and this led me 

 to fresh exertions, especially with a view to carry the elevation 

 of temperature to a greater pitch than I had hitherto done, 



" If we employ spectrum tubes in which the gas is extremely 

 rarefied, little is gained in the way of increasing the luminosity 

 of the spectrum by pushing, beyond a certain limit, the power 

 of the induction coil, whose discharge traverses the tube ; but 

 by operating in the manner indicated in a former memoir, a 

 new course is opened to us. M. Hittorf, Professor of Chemis- 

 try and Physics in the University of Munster, was kind enough 

 to associate himself with my recent labours ; and in confining 

 myself at this moment to one class of phenomena, I shall select 

 from our experiments those which illustrate the transformations 

 experienced by the spectrum of the same gas, as its temperature 

 is augmented more and more by the passage of currents of in- 

 creased strength. In the first place I will allude to the spec- 

 trum of hydrogen. If we pass the discharge of a large Ruhm- 

 korff coil through a capillary tube, very narrow and not long, 

 filled with this gas at a pressure of about half an atmosphere, a 

 spectrum is obtained similar to that afforded by employing a 

 small coil and a great rarefaction of gas ; but if we interpose, 

 as suggested by M. Ruhmkorff, a Leyden jar, to increase the 

 energy of the current, the spectrum completely changes its 



