128 REPORTS ON THE STATE OF SCIENCE. 
the one they show in the electric arc. The difference will be seen from 
the following diagram (fig. 5):— 
This shows the spectrum of barium when it is used as metallic 
anode in a Wehnelt tube, and when the currents sent through the gas 
are so great that the metal begins to evaporate. Then 
the light near the anode throws a spectrum which is 
about the same as the one the metal sHows in the 
electric arc. Below this spectrum is shown that of 
barium anode rays. It is easily seen that the first 
spectrum has many more lines than the second one, 
which is the spectrum of the anode rays. 
The barium rays show a spectrum of the same type 
as that of the rays due to other alkaline earths, as 
they all have one line, which has no relation to any 
other, an isolated line, and then a few pairs of the same 
difference of vibration-frequency which are connected 
with the atomic weight in the well-known way detected 
by Runge and Paschen. So we may suppose that the 
lines owe their origin to charged atoms. 
By studying the spectra of anode-rays it will be 
possible to decide also for other elements which lines 
are emitted by isolated atoms and which are due to more 
complex systems. And as these spectra are very simple 
we may hope that relations in those spectra will be found 
for which no such relations have yet been discovered till now: for 
instance, in the spectrum of iron; but my experiments in this direction 
are not yet sufficiently advanced to say anything definite about it. 
All lines of anode rays which were studied are found to give the 
Doppler effect; by means of this effect we measured, for instance, the 
velocity and ratio of e by m of sodium rays, and found a mass equal 
to that of the sodium atom. 
I may mention that, when canal rays are moving towards the 
observer, one finds two spectrum lines instead of one. One of these has 
the ordinary wave-length and one is shifted towards the violet ; in Ger- 
man called by Stark ‘die ruhende Intensitaét ’ and ‘ die bewegte In- 
tensitit.’ 
The explanation of the shifted line is given by the theory of the 
Doppler effect; but the origin of the ‘ ruhende Intensitat’ is prob- 
lematic, and many and not very satisfactory theories are made about it. 
For the solution of the problem of the unshifted line a result may be 
important which I found in the Doppler effect of the rays from the 
alkaline earths. The spectrum of Sr-rays shows one isolated line and 
two pairs. The surface of the anode emits only the isolated line; the 
other lines only appear in the ray itself. And in observing the Doppler 
effect one finds that only the isolated line shows both the shifted and 
unshifted lines; the first being emitted by the ray, the second by the 
surface of the anode. The pairs only show the shifted lines, and not 
the unshifted. 
In the case of canal rays there are always the same lines emitted by 
the ray and the surrounding gas, so it seems to be quite natural to find 
both the shifted and unshifted lines. 
Here, in the case of anode-rays, the ray emits lines different from 
those of the gas, and so one finds, as to be expected, only the shifted lines. 
Fic. 5. 
