174 ANNUAL OF SCIENTIFIC DISCOVERY. 



well-known mottled appearance which the surface of the sun exhib- 

 its is due to the presence of " willow-leaf-shaped " luminous bodies, 

 which, interlacing, as it were, cover the whole surface of the sun. 

 These most singular forms can be well observed, according to Mr. 

 Nasmyth, in the " bridges " or streaks of light which cross the dark 

 spots, and they are there seen to move with an astonishing velocity. 

 Imagination itself fails to give us the slightest clue to the probable 

 constitution of these most recent of astronomical novelties ! 



The beautiful red prominences seen projecting from the sun's disc 

 during a total solar eclipse, and reaching to a height of forty thou- 

 sand miles above the sun's visible surface, are likewise objects whose 

 existence cannot be reconciled with any of the proposed theories 

 of the sun's structure.- Edinburgh Review. 



Mitscherlich's Researches on Spectral Analysis. In examining 

 with the spectroscope a substance containing baryta, M. Mitscherlich 

 observed two bright green bands which appeared to indicate the pres- 

 ence of a new metal. On further investigation it was found that the 

 same lines were obtained, sometimes alone, sometimes together with the 

 barium lines, when a solution of chloride of barium containing sal- 

 ammoniac Is employed. Acting upon the data thus obtained, M. Mits- 

 cherlich continued his investigations, and has been led to the conclu- 

 sion that metals do not give a spectrum in all their compounds, and 

 that they do not give the same spectrum in different compounds, 

 but that the character of the spectrum depends upon whether it is 

 produced by the metal, or by one of its compounds of the first order. 

 It further appears that every compound of the first order, if it have a 

 spectrum other than that produced by decompositon, must have a 

 spectrum of its own. Metallic compounds are so easily reduced by 

 the flame that we usually obtain only the spectra of 'the metals them- 

 selves. Light passed through ignited soda vapors, or vapors of the 

 carbonate of soda, does not give the sodium line ; but the vapor of 

 metallic sodium, at a low red heat, exhibits this line distinctly. 

 From this it follows that in those flames which exhibit the sodium line, 

 metallic sodium, as such, produces the line in question, and since so- 

 dium has almost the greatest affinity for oxygen, it follows that all spec- 

 tra which are produced by oxides are metallic spectra. The author 

 further suggests that these experiments enable us to determine the 

 affinities of the elements at the temperature of the sun's atmosphere 

 by the spectral analysis. If, for instance, we observe the spectrum of 

 a particular metallic chloride in the sun's light, we should have to con- 

 clude that at the temperature of the sun's atmosphere the metal in 

 question has a greater affinity for chlorine than potassium or sodium, 

 since these exist as metals in the sun's atmosphere. Moreover, we 

 may hereafter, conversely, determine the temperature of tho sun's 

 atmosphere from the nature of the chemical compounds which exist 

 in it. provided that we succeed in obtaining an approximately high 

 temperature. 



From the fact that free potassium and sodium exist in the sun's 

 atmosphere, it follows that no free electro-negative body like oxy- 

 gen, sulphur, etc., can be present, and not even enough to combine 

 with all the sodium. Consequently all metals which are reduced 

 from* their compounds by sodium must exist in the sun's atmosphere 



