1882.] on Spectrum Analysis in the infra red of the Spectrum. 



61 



was rotating ; and developed before the audience and subsequently 

 the photograph was shown.] The exposure is estimated by Cazin 

 as TTj^o^oo" ^^ ^ second. It would require twenty such sparks to 

 impress the red end of the spectrum on a pure bromide plate. I 

 should wish to show you one more remarkable example of the 

 action of the dark rays on the blue-green silver bromide. On the 

 screen we have a slide showing certain opaque discs and triangles. 

 They were produced in the following manner : A card was perforated 

 with such discs and triangles, and placed i of an inch above a green 

 bromide plate ; above this was suspended a kettle of boiling water, 

 and the radiation from the kettle acted on the plate through these 

 holes, with the result that, after considerable exposure, an image 

 of the holes was developed on the bromide below. This shows 

 that this particular form of silver salt responds to waves of very low 

 refrangibility. 



The first application of the new compound was to the solar 

 spectrum, and on the screen we have the first impression of the infra- 

 red region ever taken. On the diagram (Fig. 2) there are bands and ij/ 

 drawn which do not appear in the photograph ; only on two occasions 

 have they been impressed, for reasons which wdll be explained. To 

 show you how far our knowledge of this region is extended, a 

 photograph is shown on the screen of the spectrum obtained photo- 

 graphically by Draper, which he obtained by indirect means. 



Fig. 3. 



When a grating of large dispersion replaces the prism, the bands 

 are broken up into lines ; and very beautiful lines they are in some 

 cases. From such photographs a wave-length map was made. 

 [Shown]. The line of greatest wave-length impressed in the spec- 

 trum is 22,000. Now the visible part of the spectrum extends from 

 X 3800 to A 7600 ; thus the invisible sj)ectrum, as photographed, is 

 five times longer than the visible spectrum. 



In the visible portion of the solar spectrum, most of the lines 

 have been traced to the absorption of different metallic or other 

 vapours existent in the solar or our own atmosphere. The cause of 

 the absorptions in the invisible part of the spectrum are as yet 

 untraced, except in one or two instances to which I shall have to 

 allude presently. With the exception of sodium and calcium, no 

 metallic vapours seem to have what would be bright lines, were 

 they visible, in the infra-red portion of the spectrum ; hence we are 



