60 



connection witli meteorological changes, cannot be explained 

 by the hypotbesis tbat it is due to dust. The balo or corona 

 round tbe sun is very frequently visible now, and has been 

 before tbe date of tbe glows. 



In 1880, in connection witb a set of inquiries I bad resolved 

 to mal^e into tbe zodiacal ligbt, I bad been paying particular 

 attention to tbe appearance of tbe sun's surroundings, and I 

 find tbat many items of tbe following notes agree closely witb 

 tbose relating to sky appearances near tbe sun in 1880 and at 

 •otber times doubtless : — Tbe sky rarely bas tbe same tint over 

 its wbole surface, tbe blue tints being about midway from sun 

 to borizon, approacbing tbe sun and borizon at certain times 

 ^nd states of weatber. Tbe cbauge of tint near tbe sun forms 

 a balo. Tbis changes in form, it is not always circular, in fact it 

 is rarely circular. Its limb bas various breadths. It always 

 changes as the sun approaches the borizon. It changes in tint 

 as the sun sinks and rises. As the change from sunrise to 

 noon is similar to sunset but in somewhat reverse order, I shall 

 only describe tbe change as tbe sun sinks. As the sun sinks 

 tbe central area becomes apparently brighter. Tbe breadth of 

 tbe limb increases on tbe side nearest to the horizon. It then 

 appears pear-shaped towards sunset. The central space 

 lengthens from tbe sun on tbe side farthest from tbe horizon. 

 Tbe rate of change is not constant, increasing as the sun sinks. 

 The tint of tbe balo is not the same at tbe same time on 

 Tarious dates. It varies from a salmon tint to a panther tint, 

 xind frequently it bas different tints in different parts. Tbe 

 panther tint occurs when somewhat heavy clouds are present ob- 

 scuring portions of the balo, which is then much more extended. 

 When it is tbis colour tbe glow is not prominent. Tbe rain- 

 bands low down are then heavy. Tbe glow varies with the 

 balo. As tbe sun still sinks this pear-shaped surrounding 

 divides at the apex, forming two dusky pyramidal columns whose 

 base rapidly extends north and south along the borizon, 

 changing in tint at different parts. These bases do not travel 

 north and south to like extent. Seen from various parts in 

 tbis colony they generaUy seem to travel farthest north. As 

 the sun sinks below the borizon tbe brilliancy of the central 

 part becomes apparent directly, so tbat it appears to increase 

 in brightness. Its tint then varies from a steel colour to 

 yellowish white on various days. This central portion seems 

 to extend to tbe point occupied by tbe limb before it divided. 

 AVhen tbe balo is extended the glow lasts a long time. When 

 salmon-tinted tbe glow is lengthy. When there is no halo 

 there is no glow proper, but a yellowish or orange bar followed 

 by a whitish shimmer. There have been a few splendid 

 ^greenish twilights in form and duration nearly equal to tbe 



