410 



Dr. B. Stewart. 



[Mar. 9, 



balanced by that of sodium at the temperature of the flame of cya- 

 nogen burning in air, but is sensibly less than that of sodium, at the 

 temperature of a jet of coal-gas and oxygen, much less than that of 

 sodium in the oxyhydrogen jet. This seems to render it probable 

 that the temperature of the incandescent thread is not far different 

 from that of a cyanogen flame burning in air (or rather the tem- 

 perature it conveys to the sodium in it), but is Jess than that 

 of an oxyhydrogen flame, though this does not necessarily follow 

 from the experiments, inasmuch as the radiation of the sodium is so 

 much more limited as to range than that of the carbon. When a 

 Bunsen burner or a gas blowpipe flame was interposed between the 

 lens and slit, no reversal of the hydrocarbon bands could be seen. 

 When magnesium was burnt between the lens and slit, the magne- 

 sium lines (o) were seen bright, eclipsing the carbon. Possibly the 

 smoke of magnesia may have considerably helped to eclipse the light 

 of the carbon. 



IV. " Preliminary Report to the Solar Physics Committee on a 

 Comparison for Two Years between the Diurnal Ranges of 

 Magnetic Declination as recorded at the Kew Observatory, 

 and the Diurnal Ranges of Atmospheric Temperature as 

 recorded at the Observatories of Stonyhurst, Kew, and 

 Falmouth." By Balfour Stewart, LL.D., F.R.S., Pro- 

 fessor of Physics at Owens College, Manchester. Com- 

 municated to the Royal Society by permission of the Solar 

 Physics Committee. Received January 25, 1882. 



1. In a paper communicated to this Society, and published in its 

 "Proceedings" (vol. 32, p. 406), evidence was brought forward tend- 

 ing to show that what may be termed declination-range weather takes 

 1*6 days to pass from Toronto to Kew ; that is, the same phase occurs 

 1-6 days later at Kew than at Toronto. And in a previous paper (op. 

 cit., vol. 29, p. 308) evidence was brought forward tending to show 

 that temperature-range weather takes about 8 days to pass between 

 these two places. * 



In this last-mentioned paper an attempt was likewise made to show 

 that there is a similarity between magnetical and meteorological 

 changes, and that both are due to the sun. This result has been con- 

 firmed by subsequent discussion, and there seems reason to suppose 

 that in America both magnetical and meteorological changes follow 

 very quickly after the solar changes which produce them. 



