DR L. BECKER ON THE SOLAR SPECTRUM. 



13( 



dition, some of them referring to a column of oxygen even three times as long as M. Janssen 

 demands. There are, however, many faint telluric lines within the space A = 580 to 572 

 which are visible at medium altitudes, when the atmosphere traversed equals only one- 

 fifth of the length M. Janssen considers essential for their visibility. We are inclined 

 to believe this to be due to the excellent optical appliances at our disposal, as shown by 

 the great number of faint lines now observed for the first time. 



However, there still remain numerous faint lines between this band and the rain-band, 

 which, although of the same order of intensity as those of M. Janssen's oxygen-band, 

 could hardly be grouped with them. Nor could they belong to the rain-band, a few dark 

 lines excepted. Before beginning this work we were struck by a relation between the 

 oscillation frequencies of the head lines of the groups A, B and a and of the region 

 under consideration, which Professor Piazzi Smyth named in his maps Region of " Low 

 Sun Band " of Thin and Closely -set Telluric Dry -Gas Lines. If we suppose the relation 

 between the oscillation frequencies of A, B and a not to be accidental, we should expect 

 an oxygen-band to end at A = 5788 A.U., or very near to the place where M. Janssen's 

 band begins. This supposed oxygen band could not be identical with the latter, because, 

 according to M. Janssen, the A, B and a groups increase proportionally to the first power 

 of the density. Now our observations give a band of closely-set lines, which ends at X = 

 5788, and which does not at all present the general aspect of the water- vapour bands. It 

 still remains an open question whether this is the result of a fortuitous coincidence, or is 

 due to some unknown law. 



The two isolated groups of lines from A = 5538 to 5386 and from 5111 to 4981 

 are produced by a variable element of our atmosphere. We conclude this from the 

 behaviour of the darker lines only. 



We have now to draw attention to a remarkable relation among the oscillation fre- 

 quencies which correspond to the middle of the water-vapour groups. Attributing the 

 two isolated groups of lines in the green and green-blue to the absorption of water-vapour, 

 we obtain the following values of i , each being the mean of — of the first and last prominent 

 lines of each band. The values of the inverted wave-lengths of the water- vapour bands 

 in the red end of the spectrum have been taken from Professor Smyth's maps and my 

 own observations in 1889. 



a, .... 



Water- vapour group near B, 

 „ „ near C, 



Rain-band, 



8-band, .... 



Water-vapour group A = 5538 to 5386 (£), 



„ A = 5111 to 4981 (t), 



1 



A 



No. of Band 



1380 



1 



(1434) 

 1533 



2 

 3 



1684 



4 



(1748) 

 1833 



5 



6 



1978 



7 



The inverted wave-lengths of the first, third, fourth, sixth, and seventh bands form very 



