124 
ME. W. GEOFFREY DUFFIELD ON THE 
Of the results obtained by Humphreys and Mohler, the following are tabulated 
as being pertinent to the subject of this paper :— 
(1) The Shift is proportional to the excess of pressure above one atmosphere, and 
takes place towards the red end of the spectrum. 
(2) The Shift is different for different elements. 
(3) The Shift differs for different groups of lines in the spectrum of any one 
element. 
(4) The Shift of lines belonging to the Second Subordinate Series is twice the shift 
of those belonging to the First Subordinate Series, which is itself twice that ' 
of those belonging to the Principal Series, i.e., the shifts are in the ratio 
4:2:1. 
(5) The Shift of some iron lines is three times the shift of other lines in the same 
Spectrum. 
(6) The Shift is proportional to the wave-length for lines of the same series. 
(7) A few lines showed no broadening. 
The effect of pressure on the spark discharge in liquids and gases has now been 
investigated through a wide range by Hale, # Hale and Kent,! Lockyer| and 
Anderson, § and it has been shown that the effect of self-induction and capacity in 
the spark-circuit also affects the frequency of the vibrations ; it is not, therefore, 
possible to compare quantitatively the results of their investigations with those of 
Humphreys and Mohler, who used a direct-current arc. 
2. The Apparatus. 
The pressure cylinder, Plate 1, figs. 1 and 2, was designed by Mr. .J. E. Petayel, 
F.Pv.S., and constructed by Mr. Chas. W. Cook, of the Manchester University 
Engineering Works. This cylinder was used by Hutton and Petayel for their 
‘ Preliminary Note on the Effect of Pressure on Arc Spectra,' to which reference 
has already been made. 
The pressure apparatus consists of a cylinder of drawn steel, fig. 8, 2 feet long, 
3 inches internal and 5 inches external diameter, lined inside with brass inch 
thick ; to the top and bottom heavy flanges are screwed, and to these are bolted the 
covers which carry the arrangement for feeding the arc mechanically. The rods 
passing through the covers are screwed where they pass through the insulated plates 
at the top and bottom of the apparatus, and these are operated by the hand-wheels 
shown in the diagram. The current passes through the feed-rods to the electrodes 
which are attached to their extremities, and which form an electric arc opposite a 
* Hale, ‘ Astrophysical Journal,’ XV., 132, 1902. 
f Hale and Kent, ‘Astrophysical Journal,’ XVII., 154, 1903; and ‘Publications of Yerkes 
Observatory,’ Vol. III., Part II., 1907. 
+ Lockyer, ‘Roy. Soc. Proc.,’ LXX., 31. 
§ Anderson, ‘Astrophysical Journal,’ XXIV., 221, 1900. 
