498 Mr. J. Barnes on the Analysis of 
the mercury was placed in an atmosphere of hydrogen tlie 
characteristics of the visibility curves were not changed. My 
results show, however, that when mercury is placed in the 
presence of air both in the vacuum-tube discharge and in 
the arc, which will be described later, the appearance of 
the interference-bands is clearly changed, which can only be 
due to a change in the oscillations of the atomic systems. 
Schuster, in a lecture at the Royal Institution in 1881, drew 
from his results the conclusion: ‘ Placing a molecule in an 
atmosphere of a ditterent kind—without change of tempera- 
ture—produces the same effect as would be observed in 
lowering the temperature.” In a note to the Astrophysical 
Journal* he says “Something similar seems to take place 
as regards pressure for the sodium lines may be obtained 
wide or narrow according as the atmosphere producing the 
pressure consists of sodium molecules only or of molecules 
of a different nature.” The results here obtained seem to 
corroborate those of Schuster. 
As being of some importance in this subject [ have intro- 
duced figs. 14 and 15 (Plate X XVI). showing the broad bands 
of the sodium lines, separated and superimposed, obtained 
with a sodium flame in air as the source. With jsodium in a 
vacuum-tube these bands are as sharp as those of the mercury 
lines on Plate XXV. Fig. 13 was obtained by the green 
radiation from mercury in a tube which had been used a 
considerable time. The separation of the plates was 6 mms. 
Here not even one component is visible. A comparison of 
this photograph with that of Fabry and Perot reproduced in 
the Astrophysical Journal, May 1901, may interest the 
reader. This reproduction is of the fringes of the same line 
with the same separation of plates, but shows the components. 
Figs. 13, 14, 15 (Plate X XVI.) have been magnified about 
five times. Fig. 16 has not been magnified, and shows how 
sharp the bands are when the plates are separated 1 cm. 
Here also the components of the mercury green radiation are 
invisible. 
With tubes containing capillaries whose diameters are 
greater than 2 mms. the light obtained with an ordinary dis- 
charge is not sufficiently intense to show the finer components. 
The components that can be seen have their edges quite 
sharp, showing that the vibrations in these tubes are probably 
the same as in the tubes of smaller capillaries. The finer 
the capillary the greater the electrical resistance to the 
discharge and hence a rise in temperature, causing a brighter 
light. Temperature is an important factor, for by heating 
* Astrophys. Journ. ili. p. 292 (1896). 
