266 Flacker on Spectrum Analysis. 



vapours of iodine and bromine, offer examples under the con- 

 ditions described in my memoir. 



" If the index of refraction of two successive colours differs 

 very little, the two corresponding bands are partially super- 

 imposed the one over the other, and then, if a good telescope is 

 used, the middle of the composite band exhibits a double in- 

 tensity, sharply bounded by two bands, the breadth of which is 

 equal to half that of the slit. As the slit is narrowed more and 

 more, the most luminous central portion diminishes in breadth, 

 and disappears entirely when the breadth of the direct image of 

 the slit is less than the distance of the median lines. The two 

 simple bands are then separated by a dark space. The distance 

 of lines in the midst of the two bands* is independent of the 

 width of the slit. The beautiful double ray of mercury affords 

 an illustration. 



cc If the incident ray contains a continuous series of colours, 

 the intensity of which decreases rapidly as their refrangibility 

 increases, while the colours immediately inferior in refrangibility 

 are wanting, the corresponding portion of the spectrum pre- 

 sents a space which is very luminous towards the red side, and 

 becomes more and more obscure towards the violet. If similar 

 spaces succeed each other, the appearance is presented of a 

 column grooved and illuminated by daylight. The blue and 

 violet portions of the spectrum of nitrogen, when seen through 

 a good telescope, behave in this way. Analogous results occur 

 in the case of spaces whose illumination diminishes from the 

 violet to the red, of which I will hereafter cite an example. If 

 the incident light contains, within certain limits, continuous 

 colours, with the exception of periodical interruptions, the spec- 

 trum obtained is divided by double lines into a series of coloured 

 spaces. I have counted in the red, orange, and yellow portion 

 of the nitrogen spectrum eighteen coloured spaces, all of the 

 same breadth. If the spectrum is fine, we observe two of these 

 spaces added to the shadow next the yellow, and three in the 

 green that follows it. The spectrum of sulphur, jvhich M. 

 Geissler was the first to obtain, is entirely composed of similar 

 coloured spaces, the breadth of which is augmented from the 

 red to the opposite side. 



" If we admit that the active force developed by the heat that 

 renders the gas incandescent is of the same order in the case of 

 a continuous spectrum, as in a spectrum composed of one or 

 many bands of homogeneous light, we must conclude that the 

 intensity of these bands is infinitely greater than that of light 

 of equal refrangibility in the continuous spectrum. It follows, 

 on one hand, that we must reject all idea of absorption to ex- 

 plain the appearance of similar bands in the place of a continuous 



* " Milieu des deux bandes." 



