Colours. 211 



bow regularly diminishes. Lastly, when the sun 

 is 42 above the horizon, the axis being sunk the 

 same number of degrees below that circle, the 

 summit of the rainbow touches the horizon : so 

 that, when the sun is higher than this no primary 

 bow can be seen. A portion, however, of the 

 exterior or secondary bow, may be seen, if the 

 sun have any elevation between 42 and 54. 



If we stand on an eminence, when the sun is 

 at the horizon, a rainbow exceeding a semicircle, 

 (and, indeed, in favourable circumstances, ap- 

 proaching to an entire circle), may be seen, 



As the cause of colours must be now apparent 

 to you, and as it is evident that they must pro- 

 ceed from some quality in bodies or their surfaces, 

 which causes them to reflect rays of a particular 

 hue, you will easily understand why some bodies, 

 which are called semipellucid, afford one colour 

 by transmitted, and another by reflected light. 

 The truth is, the beam of light in passing through 

 them is dissected and separated, and part of one 

 colour is permitted to pass through, and part is 

 sent back. If a solution of a wood called lignum 

 nephriticum is put into a clear phial, when viewed 

 only by the reflected light which falls upon it, the 

 solution will appear blue ; but if held up against 

 the light, and seen through, the colour will be a 

 fine yellow. The same is found to be the case 

 with some precious stones, and some glass compo- 

 sitions. Thus, if a small quantity of arsenic is 

 mixed in the composition of glass, the mass will 



