480 
DR. WALTER NOEL HARTLEY ON THE ABSORPTION SPECTRA 
In the examination of other aromatic hydrocarbons the bronze tube was used, and 
such differences do not occur. 
On the tabulated statement the wave-lengths of the bands measured by Patter, 
Friederichs, and Grebe, are quoted for comparison. 
The accuracy of the measurements compares favourably with those made by 
Friederichs and by Grebe, who employed a small concave grating of 1 metre radius, 
because, as I have previously shown, the separation of the lines of different 
refrangibilities, with a quartz spectrograph having only one Cornu’s prism of 60° C., is 
very great in the ultra-violet, when the quartz lenses are not achromatised. The 
standard lines of reference were the wave-lengths of the cadmium spark spectrum 
determined by Eder and Valenta. 
There is evidence that the differences in the measurements of the bands in the 
different spectra observed by them are not due to inaccuracies, but to actual differences 
in the positions, or in the width of the bands ; this arises from the different 
temperatures and pressures at which the several observations were made. Much 
more careful attention to atmospheric pressure and temperature becomes necessary 
when prolonged exposures extending from 1 \ to 24 hours are made than when the 
period is only two minutes, but even under the latter conditions it is important. 
On examining the various plates of spectra it was at once observed that, as the 
temperature rises, the bands which lie at the less refrangible end between XX 2650 
and 2700 (see Plates 32 and 33, figs. 1 and 5) become sharper and more intense ; at the 
same time the bands in the more refrangible rays between XX 2300 and 2500 become 
feeble, but this is because they are merged in a general absorption which weakens the 
emission spectrum. Pairs of bands in the less refrangible rays, which are so weak as 
to be barely visible at 25° C., become well defined and strong at 43° C. and 53° C. 
For examples, take the following. The measurements are recorded on each side of 
the bands when they are distinctly visible ; the stronger edge of a group of bands is 
indicated by the thickened line of the angle. 
At t 
A band visible. 
Only just visible. 
12°‘7 C. 
/2670 
| 2672 
2631 
At t 
A band, sharp and strong. 
A faint band. 
A band just visible. 
25° C. 
/2670 
/2661 
/2625 
| 2672 
I 2664 
| 2631 
