( 896 ) 
line, and, applying DorPrer’s principle, would conclude that the 
radiating vapour moves toward the observer. 
Properly speaking, we are not allowed to apply the formulae (15) 
and (16) to all waves between d, and d,, for where x is not equal 
to zero, equation (14) should be used. Let us suppose that only in 
the middle part of that region x has appreciable values; then we 
probably shall obtain a fairly true intensity curve, when subtracting 
the ordinates of an absorption curve (supposed to be symmetrical with 
respect to OO) from the ordinates of the displaced, unsymmetrical 
curve d, d, d,. The result is a sharp drop in the intensity curve, 
representing a narrow dark line in the spectrum, whose “centre of 
gravity” is somewhat displaced toward greater wave-lengths (with 
respect to 4,), and which is partly caused by absorption, partly by 
scattering. If one should mistake this line fora mere absorption line, 
its displacement toward the red would make one think, that the 
absorbing vapour recedes from the observer. *) 
The above particulars which, according to our theory, the dis- 
tribution of the light in a wide dispersion band must show, bear a 
striking resemblance to the phenomena really observed by CHarLes 
E. Sr. Joun’) in the calcium lines /7 and A’ of the solar spectrum. 
And if, besides the consequences of anomalous scattering, we also 
consider those of anomalous refraction (not noticed in this paper), the 
agreement between the results of theory and of observation proves 
to extend to almost every detail of the phenomena described by 
ST. JonN. So it is possible to explain the rather intricate pecularities 
exhibited by the components /7,, A, 7,, K,, A,, A, of the well-known 
broad calcium lines in the spectrum of the various parts of the soiar 
disk, without having to admit with Sr. Jonn, that there is a general 
radial circulation of the calcium vapour going on in the solar atmos- 
phere, with velocities that would amount to 1,97 kilometer per 
second in the mean for the ascending, and to 1,14 kilometer per 
second in the mean for the descending motion. I must refer the 
fuller discussion of these interesting observations to a subsequent paper. 
With most lines of the solar spectrum the total region of the 
dispersion anomaly, from d, to d,, is so narrow, that the particulars 
concerning the part included between d, and d, escape our obser- 
vation. What then remains visible, is only the asymmetry of the 
1) In a former communication (Proc. Roy. Acad. Amst. XIII, p. 10; Astroph. 
Journ. 31, p 428, 1910) I wrote that the central part of the K-line, the true 
absorption line, cannot be displaced by anomalous dispersion. | did not yet realize, 
at that time, that even the central line might be an impure absorption ine. 
2) Crarres EK. Sr. Jonny, The general circulation of the mean and high-level 
calcium vapor in the solar atmosphere. Astrophysical Journ. 32, p. 38—82 (1910). 
