91 



from greatest to least, will become at once manifest by mere inspec- 

 tion. Lastly, a table is given containing a series of substances ar- 

 ranged according to their refractive powers, chiefly deduced from ex- 

 periments made according to the method here described ; some to 

 which the machine for measurement would not apply, being ob- 

 tained by other means, or borrowed from other authors. 



The second part of the paper treats of the dispersion of light. The 

 principles and observations on which the inductions here given chiefly 

 depend, are these : When a glass prism is placed in contact with 

 water, and brought near the eye, in such a position that it reflects 

 the light from the window, the extent of perfect reflection is seen to 

 be bounded by a fringe of the prismatic colours in the order of their 

 refrangibility. But it may happen that two media which refract un- 

 equally at the same incidence, may disperse equally at that incidence ; 

 and, under these circumstances, a pencil of rays passing from one of 

 these media into the other, will be refracted without dispersion of its 

 colours. The boundary of prismatic reflection will then be found a 

 well defined line, free from colour, if the surface at which the reflected 

 light emerges from the prism be at right angles to its course. More- 

 over, when the disparity of the dispersive powers of the media is still 

 greater, it may also happen that the usual order of prismatic colours 

 will be reversed, and then the red, or least refrangible ray, will appear 

 strongest and lowest in the fringe, unless the colours so produced 

 are counteracted by refraction at their emergence from the prism. 



This doctrine is illustrated by examples of various, both simple and 

 compound, substances, and especially by the effect of metallic solu- 

 tions differently diluted in less dispersive media. Having compared 

 several of these, each diluted till the limit of reflection appeared void 

 of colour when in contact with a rectangular piece of plate-glass, he 

 deduced thence a table of their refractive powers, in that state of di- 

 lution in which the eye could discern the disappearance of colour. 



He likewise made experiments on dispersion by means of wedges, 

 in the manner practised by Mr. Dollond, Dr. Blair, and others ; and 

 has reduced the substances thus examined into a second table, ar- 

 ranged according to the excess of their effect on violet above red 

 light, at a given angle of deviation. By comparing this with the 

 preceding table, it appears how little correspondence there is between 

 them, and consequently how numerous are the combinations by means 

 of which a pencil of rays that passes through two media, may be made 

 to deviate without dispersion of its colours. 



At the close of this paper the author remarks, that the colours into 

 which a beam of white light is separable by refraction, appear to him 

 to be neither seven, as they are usually seen in the rainbow, nor re- 

 ducible to three, as some persons have conceived ; but that by em- 

 ploying a very narrow pencil of light, four primary divisions of the 

 prismatic spectrum may be seen, with a degree of distinctness which, 

 he believes, has never been described nor observed before. These 

 colours are red, yellowish green, blue, and violet, in the proportion 

 nearly as the numbers 16, 23, 36, 25. 



