316 Prof. R. W. Wood on the Dispersion of Sodium 
they could be made distinct again and the count continued. 
In this way it has been possible to record shifts as high as, 
400 fringe-widths. Beyond this point it is impossible to go 
with the helium light, the components of helium ‘light: 
becoming too much spread out by the dispersion of the. 
vapour to be brought together again by going out any farther 
in the fringe system. To extend the observations to higher. 
temperatures and greater vapour-density, the green radiation 
from a mercury-tube was used, with which fringe-shifts as 
high as 100 fringe-widths were observed, corresponding to a 
shift of 2500 helium fringes. In this way it has been pos- 
sible to obtain measures of the optical density of the vapour 
under the specified conditions from a temperature of 280°, at 
which point the fringes commenced to move, to above 650°, 
at which temperature the dense vapour was luminous, the 
entire section of the tube appearing filled with a dull red 
glow. Plotting the fringe-shifts as abscissee and the tem- 
peratures as ordinates, we obtain curves of form similar to the 
density-curve obtained by Jewett, though by no means 
coincident with it, for reasons which I have already given. 
The observations were made both when heating and cooling. 
At first it was found that the shift obtained on cooling was 
greater than the one observed during the heating; in other 
words, the fringe system did not return to the starting-point. 
This was found to be due to hydrogen gas liberated from the 
sodium, which, from the circumstance that it shifts the 
fringes in the opposite direction from the shift produced by 
the sodium vapour, retards the motion of the system during 
its liberation, and by remaining unabsorbed in the tube pre- 
vents the fringes from returning to the starting-point. This 
trouble was eliminated by heating the tube to a high tem- 
perature and setting the mercurial pump in operation. The 
observations extended over several weeks, many tubes both 
of steel and glass being used. The length heated was varied 
and the different series of observations compared. Most of. 
this work was only useful in determining and eliminating 
the sources of error. 
The final series, which was considered the best, was made 
with a tube of Jena glass, the length of the heated portion 
being 8 cms. The retardation, or rather the acceleration, 
since we are working with light on the blue side of the ab- 
sorption-band, is due to a column 16 ems. long, since it is 
twice traversed by the interfering beam. From the data 
given in this table the refractive index of the vapour for 
either helium light or the green mercury radiation can 
be calculated for any temperature by means of. the 
