22 
ASTRONOMY: C. E. ST. JOHN 
originator and chief supporter, announced a deduction that offers an 
opportunity of making such a quantitative test, namely: the mutual 
influence of Fraunhofer Hues upon each other.^ The observational data 
available for a crucial test are available,^ and consist of the displacements 
of 506 Fraunhofer lines at the edges of the penumbra of eccentrically 
located sun-spots observed by me. In such observations the slit of the 
spectrograph is parallel to the radius of the solar image passing through 
a spot whose position for best observation is between the limb and the 
midpoint of the radius. The displacements are a maximum at the 
peripheral edges of the penumbra. At the edge directed toward the 
limb they are in general in the direction of longer wave-length, while 
at the edge directed toward the center of the disk they are in the direction 
of shorter wave-length. These displacements were interpreted by 
Evershed and myself as effects of motion in the line-of-sight, but are 
considered by Julius to be due to anomalous refraction. 
In this phenomenon the particular form in which mutual influence 
would manifest itself is that a weak hne on the violet side of and near to 
a stronger line is displaced less, but, if on the red side, more than the 
average amount. JuHus, in making the announcement of the deduction 
to which reference is made, gave a discussion based upon a portion of the 
available data and obtained a result apparently confirming the deduction 
and considered by him a veritable proof of the theory. In fact, the de- 
duction follows so directly from the theory that its confirmation would 
greatly strengthen the position of the theory as a working hypothesis, 
while a failure of confirmation would discredit it. The questions raised 
by Julius are elsewhere considered by me in detail.^ 
These displacements are well suited for a definitive test of the theory, 
as they are purely differential and free from possible observational bias, 
since they were made without any knowledge of what they are now used 
to prove. Moreover, a large number of other lines present upon the 
same plates are available for standards of reference. It is evident that 
much depends upon the derivation of these standards. 
The 506 lines in my original paper do not form a homogeneous series 
of observations. From this it follows that standard displacements for 
different spectral regions cannot be determined with high precision by 
deducing them for each region by any smoothing-out process involv- 
ing the results for all the segregated regions. There are six series: 
(1) X 3624 - X 3724; (2) X 3879 - X 4410; (3) X 4634 - X 4829; 
(4) X 5123 - X 5349; (5) X 5598 - X 6065; (6) X 6393 - X 6643. 
The measurements represent with high precision the relative displace- 
