a 
2 ees Se ae a ii 
ry 

Bright Spectrum Lines. -.? OE 
Thallium.—A piece of metallic thallium was placed on the 
end of a platinum wire and held in a Bunsen flame. The only 
bright radiation was that of the green line, 5439. A doubling 
of bands occurred when the plates were separated only a few 
millimetres. With a vacuum-tube radiation another com- 
ponent was found with wave-length between the principal 
and first component. 
1. Standard Component. Intensity, 1. 
2. +410 x 10°? mm. i 3/4. 
a + 0-4 2 ss 1/4. 
Hydrogen.—By the kindness of Dr. Parsons I used one of 
his tubes containing hydrogen which was specially pure, the 
pressure being 1mm. The red line easily breaks up into 
three components, one on each side of the brightest component. 
1. Standard Component. ) Intensity, 1. 
2. +0°6 x 10-° mm. < 1/4 
a. 02 % bs 1/8. 
The green line is very complex, the components are so 
numerous that observations are very difficult. 
The changes in the components, due to changes in pressure, 
size of capillary, capacity in circuit, which were examined 
principally with the mercury radiations were in some cases 
tried with the other radiations considered, and the results 
were in general the same. The above results with respect to 
the relative wave-length and intensity of the components 
under the conditions specified are collected in the following 
table (p. 502), together with the results of Michelson,and Fabry 
and Perot upon the same radiations obtained in vacuum-tubes. 
Michelson’s values are taken from the curves given in his 
paper. His method does not allow the determination as to 
whether the components have larger or shorter wave-lengths 
than the standard. The second list of values for the com- 
ponents of the mercury line, \=5461, obtained by Fabry and 
Perot, are taken from a paper by Zeeman*. 
After the many long and tedious observations, together 
with the study and elimination of the errors which may enter 
into the results due to imperfect adjustments of the apparatus, 
the author regrets that he is unable to present a more detailed 
account of the variations that occur in these component radia- 
tions or satellites as they have been called. The changes 
occur so suddenly on the least change of the surrounding 
conditions, and sometimes even when no changes apparent to 
the observer were introduced, that only qualitative results of 
a very general nature can be expressed. 
During the observations upon the sharp interference-fringes 
* Astrophys. Journ. xv. p. 218 (1902). 
