seen at Portsea on the 29th of March 1848. 441 



the circle were composed of innumerable horizontal spectra 

 overlapping each other, it would still be colourless. 



If we further trace the light that passes through two vertical 

 sides of these prisms without suffering reflexion, we shall see 

 that it emerges with no change of inclination^ but with a 

 change oi azimuth depending chiefly on the inclination of the 

 two refracting faces to each other. Now prisms of ice are 

 either three- or six-sided. In either case, the two faces through 

 which the light passes can only be either parallel or inclined 

 60° to each other : when parallel they will produce no refrac- 

 tion, and when inclined 60°, a refraction never less than 23^, 

 but sometimes a few degrees greater. Hence the common 

 halo of 46° in internal diameter, produced when these prisms 

 are turned in all directions, so as to refract in all possible 

 planes that pass through the sun. But when the prisms are 

 all parallel, as we have supposed, they can refract in only two 

 directions, so that all the light which would otherwise form 

 the halo will be collected at two opposite points of its circum- 

 ference; and if all the axes of the prisms be vertical, the only 

 portions of the halo that will be visible will be those to the 

 right and left of the sun, and on the horizontal circle passing 

 through his disc: hence the two single spectra or mock-suns, 

 which formed so striking a part of this display. Nor is there 

 any difficulty in understanding why these spectra were not 

 exactly on the circumference of the common halo (see fig. 4), 

 but rather further from the sun. This was a necessary con- 

 sequence of the plane of refraction being i?icli?ied to the axes 

 of the prisms, the effect of which was to render their refracting 

 angle viitually greater than 60°, and the refraction therefore 

 greater than 23°. In the formation of the common halo, this 

 cause must also operate in a great number of the prisms; for 

 if they lie in all directions, the refraction cannot always take 

 place in a plane perpendicular to the prism's axis. This is 

 generally overlooked in popular explanations of the theory of 

 the halo, and hence arises a discrepancy that has often puzzled 

 tlie writer, between that theory and the appearance itself; lor 

 whoever regards a common halo, will see that it is not, as 

 theoretically described, a luminous ring or band of definite 

 breadth like the rainbow, but only the boundary between a 

 less luminous and a more luminous space. The whole sky 

 outside this boundary is brighter than widiin it, though this 

 brightness rapidly diminishes from the circle outwards. If 

 we regard it as a luminous ring, therefore, only its inner edge 

 is defined, and only on this edge is there any appearance of 

 colour, viz. red, orange, yellow, and then white, with perhaps 

 a very faint tinge of green ; but beyond the yellow, or the 



