seen at Porlsea on the 29th of March 1848. 4.43 



remaining where it was, would be 15' from the inner edge; 

 which edge, however, would still be very sharply defined, for 

 the light would diminish towards it like the ordinates of some 

 such curve as C (fig. 4), so that the halo would resemble that 

 formed by placing C on a black wheel, and making it revolve 

 instead of B. Finally, when the light is not homogeneous 

 but white, and comes from a source of 30' diameter, innume- 

 rable halos are formed, each resembling that last described, but 

 all having different radii, the red least and the violet greatest. 

 To imitate this effect, therefore, take several such figures as 

 C (say seven, painted with the seven principal colours), and 

 place them on the black wheel at different distances from its 

 axle, as in fig. D. This arrangement, when revolving, would 

 give nearly the appearance of the common halo ; though to 

 give it exactly, would require an infinite number of such 

 figures, at distances and with colours, intermediate between 

 R and O, between O and Y, &c. 



In this way the difference between the colours of the halo 

 and those of the rainbow may be well shown. For it is plain 

 that if a rainbow were formed by homogeneous light coming 

 from a star, or a physical point, the bow would be a mere 

 physical li7ie. Let the light come from a source of 30' dia- 

 meter, and the homogeneous bow will be 30' broad, and its 

 intensity will vary from side to side according to the ordinates 

 of a circle. Such a bow may be seen by turning a black wheel, 

 on which are one or more equal circular discs of colour, at 

 equal distances from the axle. Both the edges of such a bow 

 will be well-defined. Now to imitate the bow formed by 

 •white light, the discs must have various colours, and be placed 

 at unequal distances from the axle, as at E (fig. 4), the imita- 

 tion being more exact, the larger the number of discs having 

 different tints and distances, provided they be all confined to 

 a ring no wider than about three and a half times their own 

 diameter (for the primary, or six times for the secondary bow). 

 In the same way, a figure or repetition of figures like A, 

 placed at equal distances from the axle of the wheel, will on 

 revolving give the appearance of the large horizontal band in 

 the phaenomena above described ; and a number of complete 

 spectra arranged as at F, will produce the effect of one of the 

 bands of large radius near the zenith in figs. 1 and 2. 



As all the effects above stated, in the case of a halo, must 

 take place equally in that of the lateral mock-suns, which are 

 in fact merely detached portions of a very intense halo, this 

 explains why they appear to have colourless tails directed 

 from the sun, and fading gradually as they recede from him. 

 'I'hese tails correspond to the outer and colourless portion of 

 a halo. 



