PHENOMENA RELATING TO THE SPECTRA OF HYDROGEN AND HELIUM. 239 
(II.) Methods of Measurement. 
For a number of reasons, measurements of the intensities of spectrum lines in the 
visible region with the thermopile or bolometer appear to be impracticable. In the 
first place, the intensity of the source must be exceedingly great if galvanometer 
deflections are to be obtained of a magnitude suitable for accurate measurement, and 
it is doubtful whether any but the brightest lines could be dealt with in this 
manner.* Moreover, it is only possible by such methods to deal with one line at a 
time, and the difficulty of maintaining sources of light constant for the duration of a 
series of measurements would, in many cases, appear to be almost insuperable. It 
would be necessary to introduce corrections for the loss of light in the spectroscope, 
and the determination of the magnitudes of these corrections would necessarily be a 
matter of considerable difficulty. Measurements have been made of the relative 
intensities of spectrum lines with the spectrophotometer. With this instrument it 
is difficult to attain any high degree of accuracy, and the range of wave-length over 
which it can be used is small unless the intrinsic intensity of the radiations is very 
great. It is, moreover, necessary to encounter the difficulty referred to above of 
maintaining the light source constant during a series of measurements. 
In a recent paper,! we have discussed the application of a method, involving the 
use of a neutral-tinted wedge, to the determination of the distribution of intensity in 
broadened spectrum lines. In this case, the problem is reduced to its simplest form 
owing to the fact that over the short range of wave-length covered in a single 
broadened line, we may consider that both the extinction-coefficient of the wedge 
and the sensibility of the photographic plate remain constant. Such a method cannot 
be applied directly to the present problem without important modifications and 
additions, for in the measurement of the relative intensities of lines of widely 
different wave-length neither of these essential conditions is even approximately 
fulfilled, and, in addition, the loss of light in the spectroscope must be considered. 
The problem dealt with in the present communication is therefore of an essentially 
different nature, but it has been found possible to evade these difficulties without 
altering the character of the method employed. 
The apparatus used in the investigation is identical with that which we have 
described previously| and consists of a spectrograph, in front of the slit of which is 
mounted a neutral-tinted glass wedge, cemented to a similar wedge of colourless 
glass so as to form a plane-parallel plate. The spectra are photographed through the 
neutral wedge, and the resulting photographs consist of lines which are dark along 
the edge corresponding to the thin end of the wedge, and which fade away towards 
* Cf. Jolly (‘Phil. Mag.,’ xxvi., p. 801, 1913). A bibliography of the subject is given by this author 
and by Ivonen and Jungjohann (‘Astrophys. Journ.,’ xxxii., p. 141, 1910). 
f ‘ Phil. Trans.,’ series A, vol. 216, p. 459. 
| Loc. cit , 
2 L 2 
