

Bright Spectrum Lines. AQT 
differences in wave-length and in intensity relative to the 
one selected :— . 
1. Standard Component. Intensity, 1. 
oe TL 10-8 nr. Af 3/4. 
35. — O09 * ad 1/4. 
4, — O04 “ a 1 
2. + 01 - si 1/8. 
6. + 0-4 ig e 1/4. 
Thus there are three components on the side toward the 
shorter wave-lengths and two toward the Jar per. 
. The violet line, 4358, is a triple having slight components 
on.each side of oe principal. 
1. Standard Component. Intensity, 1. 
2. — 05+ 107% mm. sy 1/4. 
3d. + 0-4 2 = 1/4. 
Both the yellow lines have numerous components, but they 
are of very slight intensity, so that concordant results were 
not obtainable. 
When a small amount of air was allowed to enter the 
vacuum-tube till the pressure was about 5 mms. the com- 
ponents of smal] intensity completely disappeared, the fringes 
due to the brighter components broadened and their edges 
became less sharply defined, showing that the atomic vibra- 
tions were not so uniform and simple as before. The same 
effect was noticed with the radiation from a vacuum-tube 
which had been used some time without any change of 
pressure. In the case where the pressure is changed thr ough 
the introduction of air the molecular collisions may be made 
more frequent, which would naturaily interfere with the free 
vibrations of the atomic systems and so produce a broadening 
of the bands and cause the eS intense fringes to disappear. 
In the case of an old tube, when the pressure ‘has not changed, 
there seems to be no other explanation for the observations 
than that the mercury vapour had become contaminated with 
gases driven off from the glass by the heat developed in the 
discharge. 
Whether the atomic vibrations in a source are changed on 
account of the presence of molecules of foreign matter is 
an open question. Michelson* thinks that the presence of 
other molecules does not have any appreciable effect except 
to diminish the visibility. In the case of mercury he 
obtained quite different visibility curves when the pressure 
was high to that obtained when the pressure was low. When 
* Phil. Mag. xxxiv. p. 280 (1892). 
Phil. Mag. 8. 6. Vol. 7. No. 41. May 1904. 2M 
