132 DR L. BECKER ON THE SOLAR SPECTRUM. 



Considering that there is a strong water-vapour band close to the oxygen band B, Sir 

 David Brewster's description appears to us consistent with the assumption of water-vapour 

 being the cause of the <5-band. Also, his observations on October 31 are not adverse to 

 this hypothesis, because the atmospheric lines, which did not appear so dark as usual, 

 include, besides S, other water-vapour bands in the red end of the spectrum. That at the 

 same time the rain-band had suffered a considerable absorption cannot be considered an 

 argument against this view, if we compare the number of dark lines in the rain-band and 

 <^-band in Table V. Brewster's description as well as our interpretation are ap- 

 parently in oyjposition to the observations of Angstrom, 1 who says: — "Outre les trois 

 groupes de raies situes pres deA, B et a, il existe, a gauche deD, une bande d'absorption, 

 toujours visible dans le spectre du ciel pur. Cette bande s'etend de 5681 a 5812 a peu 

 pres, et, d'apres Brewster, je la design erai daus la suite par la lettre $. Des que cette 

 bande commence a se montrer dans le spectre solaire, on peut la resoudre en raies 

 tres-fines ; mais au coucher du soleil, les raies, en se joignant, forment une bande obscure 

 et continue. Or, puisque l'apparence de cette bande ne change pas avec les circonstances 

 desquelles depend rintensite" des raies d'absorption dues a la vapeur d'eau, l'origine en 



doit etre attribute a une cause toute diff erente Pour expliquer l'origine des bandes 



A, B, a et S, qui sont tres-constantes et ne dependent pas sensiblement des variations de la 

 temperature de Fair, il faut recourir a d'autres corps gazeux moins variables en tension que 

 la vapeur d'eau." We can only reconcile Angstrom's view with Brewster's description 

 and our own observations by supposing the fainter lines within the S band to be produced 

 by dry-gas absorption. Certainly these closely set lines present a sufficiently striking 

 appearance, if viewed in a spectroscope like Angstrom's, as may be concluded from the 

 extent he gives to the <5-band towards the red. Besides, little water-vapour would 

 suffice to make the region from A =5710 to 5680, which contains so many conspicuous 

 solar lines, appear dark in a low sun. 



This hypothesis is supported by some experiments that M. Janssen 2 has made on the 

 absorption produced by oxj^gen. By employing tubes of different lengths filled with 

 oxygen under different pressures, M. Janssen discovered that several absorption bands 

 begin to appear in a tube 60 metres in length charged with oxygen under a pressure of 

 6 atmospheres, and that the same effect is produced if the length of the tube be altered 

 inversely as the square of the density. Thus he finds that the oxygen of one of these tubes 

 is equivalent to a column of oxygen 2160 metres long under the pressure of 1 atmosphere. 

 But since the oxygen of our atmosphere in the zenith equals only 172 metres at 

 normal pressure, M. Janssen concludes that the absence of these bands in the solar 

 spectrum at considerable altitudes of the sun is fully explained. The figures given in 

 Chapter 6 prove that at zenith-distances larger than 86° the band should be visible in a 

 spectroscope of the power M. Janssen employed. Most of our observations fulfil this con- 



1 Recherclies sur le spectre solaire, p. 40. 



2 Vierteljalirsschrift der astronomiscluin Geselkchaft, 25 Jahrg., 1 Heft. Leipzig, 1890. 





