442 Mr. E. L. Garbett's Description of some Parhelia 



point of maximum brilliancy, there is no other decided colour, 

 and the colourless light dies away so gradually that it is im- 

 possible to assign its outer limit. This is by no means then a 

 complete spectrum, like the rainbow. 



Now if we consider that the prisms producing the halo may, 

 owing to their various positions, produce various amounts of 

 refraction, but in no case less than a certain amount, this limit 

 being about 23°, but different for the different colours; if we 

 further remember, that the nearer a prism may be to this 

 limiting position which gives the least refraction, the more 

 slowly will the amount of refraction change with the change of 

 position ; and further, that (for more causes than one) the 

 mo7-e refraction any prism may produce, the less light will it 

 transmit ; — we shall see that if a halo were formed by homo- 

 geneous light coming from a source of no appreciable size, 

 such as a star, the circle of maximum brilliancy would be at 

 its i?ifier edge, from whence the light would diminish outwards 

 as the ordinates of some curve very concave, something like B, 

 fig. 4, which curve being cut out and placed on the blackened 

 rim of a wheel, with its broader end towards the axis, would, on 



Fig. 4. 

 being rapidly revolved, give the appearance of such a halo*. 

 But if the light came from a source whose diameter was about 

 30', such as the sun, the inner diameter of the halo would be 

 diminished by about 30', and the circle of greatest brilliancy, 



• Light figures placed on a dark wheel and made to revolve in this man- 

 ner afford a good means of illustrating a large number of optical facts 

 besides the composition of colours. To make the results obvious, the 

 figures should be repeated as many times as the circumference of the wheel 

 will contain them. 



