OF THE VAPOURS OF BENZENE AND ITS HOMOLOGUES. 
479 
cases two very narrow bands were seen so close together that the space between them 
could not be measured, and this explains why they have been assigned the same 
wave-length. In like manner, where a rather broad band ended and another 
commenced, it was found impossible to measure the two points with certainty, although 
an extremely narrow space could be seen between them. Each spectrum was 
measured twice independently, and the bands for the most part were described. For 
constructing an interpolation curve, the spectra of lead, tin, and cadmium were 
photographed with a very narrow slit and short exposure with a self-induction coil in 
circuit, by which means the spark lines were rendered sharp and narrow. But as the 
locus of the foci lies exactly in the same plane, the 18 lines of cadmium between 
Cd 17 and Cd 26 are alone sufficient, and latterly were adopted. 
The linear dispersion on the original spectra is such that between Cd 17, X 2748"68, 
and Cd 26, X 2144 - 44, there are 2781T linear divisions of the micrometer scale 
= 27011 inches, which are equivalent to a difference of 604"24 Angstrom Units 
between the above wave-lengths. The dispersion at the less refrangible end, Cd 17, 
increases from 3 "3 divisions to 6'3 divisions for 1 Angstrom Unit at the more 
refrangible end, Cd 26. 
When the wave-length measurements had been tabulated in the usual manner, it 
appeared as if the bands in the different spectra of the same substance bore no relation 
to each other ; accordingly, they were rearranged in four columns, so that obviously 
related bands or groups fell upon the same horizontal lines or thereabouts, and the 
changes in the spectra caused by increase of temperature or diminution of pressure 
are thus rendered obvious. Nevertheless, if the spectrum observed at pressure 
759'5 mm. and temperature 12°7 C. be compared with that taken at 778 mm. and 
temperature 11 0, 5 C., it will be seen that there is a considerable difference between 
them, which may roughly be indicated by the former conditions yielding 55 measurable 
bands, while the latter gives only 36. This is not to be accounted for by a difference 
merely in the quantity of benzene in the tube, as the following statement will 
show :— 
The calculated weight of benzene in 79 c.c., the capacity of the tube— 
° C. mm. grm. " 
At 11-5 and 778 = 0-0175 
„ 127 „ 759-5 = 0-0184 
„ 100 „ 757 = 0-2021 
„ 100 „ 4 = 0-0011 
The tube used for different pressures was not the same, although of the same length ; 
that for different temperatures being of glass, there may have been some reflection 
from the inner surface by which the absorption bands would be rendered less distinct, 
and this most probably is partly, if not entirely, the cause of the differences in the 
two series of spectra. 
