DISTRIBUTION OF INTENSITY IN BROADENED SPECTRUM LINES. 
485 
photographs of magnification 20'3 the breadths between the protuberances are 10'8, 
23'0, and 35*5 mm., respectively. These are in the ratios 1 : 2'13 : 3'29. The 
separation of the three closest components in Stark’s experiments* are 2'6, 6'2, and 
8'8 A.U., which are in the ratios 1 : 2'38 : 3'29. The agreement is striking. More- 
over the separation of the next component in Stark s experiments is 11*5 A.U., from 
which we deduce that the breadth at the next protuberance should be 477 mm. 
But the whole breadth of the base of the photograph is about 39 mm., and therefore 
with the degree of exposure given this component could not appear. The existence, 
therefore, of components whose separations are in the same ratio as those found by 
Stark appears to be established, and affords a strong confirmation of the view that 
the main factor which controls the broadening is the electrical resolution of the lines. 
The corresponding phenomena for are much more complicated, and the deter¬ 
mination of the attenuation-ratios and intensities of the components may, therefore, 
be deferred for subsequent treatment. 
(X.) The Diffuse Series of Helium and Lithium. 
The intensity distribution in broadened lines in the spectrum of helium has been 
investigated in the same manner as in the case of hydrogen, but the quantitative 
intensity distribution has not yet been determined. The spectrum was produced by 
passing condensed discharges through a vacuum tube containing helium with a 
spark-gap in the circuit. The results are qualitatively in accordance with the 
intensity distribution to be expected from the electrical resolution of the lines. The 
broadening of the line X = 4471 is particularly striking, and it appears to consist of 
a bright component of great intensity and a broader displaced component. Stark* 
has found that the electrical resolution of this line is unsynnnetrical and that the 
intensity of the central components is very small. Since it has been shown that the 
breadth of a component increases with its distance from the unresolved line, we should 
expect the above distribution of intensity in the line X = 4471. On the other 
hand, the line X — 4026 appears to consist of a bright central component with 
nebulous “ wings,” which is also in agreement with theory. 
It may be mentioned that by the use of the wedge method the relative differences 
in the intensities of the lines which occur when a condensed discharge or an 
uncondensed discharge is employed can at once be seen. In the case of helium this is 
particularly interesting. With the condensed discharge the two diffuse series are 
relatively much more intense than with the uncondensed discharge. It is remarkable 
also that by far the greatest increase in relative intensity is found to occur in the 
line X = 4471, whilst the D 3 line, the preceding member of the series, is affected to 
a smaller degree. This result, whilst affording no explanation, indicates the possibility 
* Log. cit. 
3 u 2 
