286 Dr. William Hiiggins [Feb. 6, 



however, Captain Abney, by the discovery of a new molecular condi- 

 tion of silver bromide, has brought the whole of the other end of the 

 spectrum, the ultra-red, within the power of the photographic plate. 

 He has, I believe, taken the photograph of a kettle of boiling water 

 in the dark by means of its own radiation. 



This evening we shall have to do exclusively with the ultra- 

 violet portion of the spectrum. 



In the years 1865 and 1869 I had the honour to bring before this 

 Institution the results of the observations of Dr. Miller and myself 

 on the visible spectra of some of the stars. These eye observations 

 embraced a range of vibrations extending from a little below C in the 

 red to about G in the blue. The recent researches, to which I now 

 at once proceed, begin where the eye observations ended, about G, 

 and carry our knowledge of the stellar spectra beyond 0, and in some 

 cases beyond S, in the ultra-violet. 



We shall, perhaps, underrate the importance of a knowledge of 

 the ultra-violet spectra of stars, if we regard these photographs as 

 simply adding so much in length to the visible spectrum, for there 

 are reasons why a knowledge of this part of the spectrum may be of 

 exceptional value to us. 



I shall describe first, in some little detail, the instrumental 

 methods by which the very great difficulties which present them- 

 selves in so delicate an inquiry were successfully overcome. The 

 two principal difficulties with which the inquirer is at once brought 

 face to face, are the feebleness of the star's light after dispersion by a 

 prism, and the circumstance that the stars are in apparent motion, 

 arising from the earth's rotation. 



It was therefore necessary to do two things, first, to obtain a 

 sufficiently pure and detailed spectrum with the least possible loss of 

 light, and secondly, to devise some method by which the star's image 

 could be kept absolutely invariarble in position within a very narrow 

 slit. 



After passing the limit of the visible spectrum, the transparency 

 of glass diminishes rapidly, until at length it becomes opaque to the 

 rays of very high refrangibility ; for this reason it was necessary to 

 avoid altogether the use of this substance. A telescope of the reflect- 

 ing form, in which the light is received upon a metallic speculum, 

 was employed. This instrument has a speculum of 18 inches diameter. 

 The spectrum apparatus must also contain no glass. There were two 

 substances available, Iceland spar and quartz, both of which are very 

 transparent to this part of the spectrum. Quartz is harder and takes 

 a higher polish and was used for the lenses, but its dispersive power 

 is so small that more than one prism would have been needed, intro- 

 ducing loss of light and other drawbacks, if this substance had been 

 employed. Iceland spar possesses a much higher dispersive power ; 

 it is, indeed, about equal to moderately dense flint glass. One prism 

 of this substance of 60°, which was beautifully cut for me by Mr. 

 Hilger, was found to be sufficient for the jDurpose. 



