154 
MR. W. GEOFFREY DUFFIELD ON THE 
approximate symmetry. The photographs also show that in all cases of un- 
. t o 
symmetrical reversal occurring between 4000 and 4500 A.U., the absorbed part of the 
line is on the violet side of the emission line. 
(This phenomenon occurring in the arc has been observed in spark discharges in 
liquids and gases under pressure by Hale and Kent,* and by Anderson.! Wilsing,| 
Lockyer, § and Hale and Kent* have discussed its bearing upon the spectra 
of New Type Stars.) 
i A 
i \ \ 
i / \ \ 
( 0 / \ \ 
I / \ \ 
> /( 2 ) \ \ 
/ .* x \ 
For unsymmetrically reversed lines under pressure the intensity curve is of the 
nature shown by the full line, fig. 11, which may be due either to the emission line 
represented by the dotted curve (l) (in which the maximum ordinate coincides with 
the centre of the absorption line) or that represented by the dotted curve (2) (in 
which it is on the red side of it), together with corresponding amounts of absorption. 
The present experiments afford a means for deciding upon the correct emission line, 
because it is not unusual to find that some lines are reversed on one photograph at a 
definite pressure, but not on another taken at the same pressure :— 
15 atmospheres 
line f 1 
is 
reversed on plate 
D25, 
but not on 
Dl9, 
20 
99 
» /2 
9 9 9 9 
D9, Dio, 
9 9 9 9 
D27, 
30 
99 
» /l 
9 9 9 9 
D9, 
9 9 9 9 
D26, 
80 
9 9 
„ c3 
9 9 9 9 
D32, 
9 9 9 9 
D15. 
The position of the most intense portion of the bright line has been determined from 
one plate, and in all cases its displacement towards the red has been found to be greater 
than that of the absorption line on the other plate, Table VII. Hence the maximum of 
the intensity curve for the emitting atom falls in such cases of unsymmetrical reversal 
on the red side of the absorption line on the photographic plate, i.e., the curve (2) 
more nearly represents the true emission line, and hence in an iron arc , for the 
* Hale and Kent, ‘ Astrophysical Journal,’ XVII., 154, 1903. 
f Anderson, ‘Astrophysical Journal,’ XXIV., 238, 1906. 
I Wilsing, ‘ Astrophysical Journal,’ X., 113, 1899. 
§ Lockyer, ‘Astrophysical Journal,’ XV., 190, 1902. 
