22 



during periods of low barometer. Under sucli conditions the 

 dust would probably be at a bigber level. The displays cer- 

 tainly seem to bave been especially fine when tbe air was dry 

 and tbe red ligbt unimpeded, during times of bigb barometer ; 

 and at such times (considering tbe downpour of air upon tbe 

 centre of an an ticy clonic system) I conclude tbat the dust 

 stratum is lower. For instance, near Warsaw, in Eussia, on 

 ]S'ovember 30tb, 1883, barometer SO 16 ; January 2nd last, 30-46 ; 

 and January 3rd, 30*27 ; on all of wbicb days tbe " red-glow" 

 is reported as being intense*. Prof. Piazzi Smytb also cor- 

 roborates tbis. So to a certain extent do my own observations 

 made at tbe Torrens Observatory, near Adelaide ; but on otber 

 occasions tbe sunset displays seem to be entirely indifferent 

 to barometric and bygrometric influences. I will give a few 

 examples. 



On January 28tb tbe barometer was bigb — 30"22 at sea- level 

 — and tbe sunset and red " after-glow" magnificent — tbe latter 

 appearing on tbe horizon about one hour and a quarter after 

 sundown like some great fire. Eelative humidity at 9 o'clock 

 was 77 i°, indicating a decidedly moist condition of tbe atmo- 

 sphere for South Australia — hardly in accord with the dry air 

 and high barometer theory so ably advocated by Professor 

 Piazzi Smyth, but on this occasion I had not access to the 

 spectroscope. 



On January 30th the same phenomena were almost identical, 

 but the barometer was loicer, relative humidity less — 69 /of — yef 

 a broad nebulous band to the left of the D line in tbe spectrum 

 indicated much moisture in the thick strata of the atmosphere. 

 Under such spectroscopic conditions I should have expected 

 that the red rays would bave been stopped off and absorbed by 

 the abundance of aqueous vapour shown by tbe spectroscope. 

 In these two instances the results of observation appear 

 decidedly conflicting. 



On March 5th the glow was again magnificent, barometer at 

 30"07 and pressure gradually decreasing, with a dry surface 

 :\i^- ;/>- .-.,1. ..r'.' .---1; '■'^, "c '0br>';.,r--^ • ■ ' ;-• 



iTiVOr- J ....... b." 



( 16th, I 



aD'-riii p ^rand dieplftj , Apoa: tiiu lelativo l.uL' 



vo;.^ low — O'l^M^ and -48}-:;; res-^- . -ano tLe spe- 



exhibited an intensely dry atmosphere, itic- iemo?i ydZoic band 

 prev^'Jled over tbe horizon, graduailj chargi-tr t^ light ora-age, 

 and ultimately becoming a biood red. Tnis ob^erration. led 



* Tlie: J baroiafeter valnes are .m<fiaDS of '" ati»ns of A3, "^Tontz'L 



^tm., ti^ke)3 at KrasniczR Wola, Grodiisk. 



