49 
of JEdinhurgh, Session 1882 - 83 . 
as a blatjk line clinging to its red side. This narrow blurred line 
represents the hundreds of thin lines produced by the absorptive 
power of the water-vapour in the atmosphere, which can be seen by 
using an instrument of sufficient • power ; and although it is never 
seen alone or measured by itself, it can be employed as a valuable 
weather indicator. 
The rainband was always observed by looking as near the horizon 
as possible, so as to take advantage of the greater depth of air seen 
through; the focus was carefully adjusted to give the sharpest possible 
definition of the solar lines, and then the intensity of the band 
together with the D-line was noted. The rainband has been 
represented by the letter tt, and the compound line, which was 
always measured, may be written shortly as D tt. 
The great difficulty in recording observations is to obtain a 
constant scale of intensities. It has usually been the practice to 
judge of the intensity of the band by the eye, and to record it on a 
scale of 1 to 5 or 1 -to 10, or else to measure it in terms of the 
low-sunband. The former plan possesses the disadvantage common 
to all purely mental scales, that there is no guarantee of the 
intensity which is represented by a certain figure one day being 
recognised as corresponding to that figure on another occasion. 
Changes in the illumination of the spectrum or in the condition of 
the observer can hardly fail to make some difference. The second 
plan is free from this defect to a great extent, but it is inconvenient, 
for beginners 'at least, because of the diurnal variations. of the low- 
sunband. Various mechanical arrangements were tried in order to 
furnish a scale which would remain constant and could be readily 
reproduced, but the results of the experiments were hot very 
satisfactory, and it therefore seemed advisable to employ a relative 
scale supplied by the spectroscope itself. The spectrum observed 
was so small that the principal Fraunhofer lines were brought near 
enough to be used as standards of comparison, and as the moisture 
in the air does not affect their relative intensities, the three most 
prominent (E, h, and F) were fixed upon as the units of the scale. 
The three increase in darkness in the order E, h, F, and they are,, 
ronghly speaking, darker in proportion to their distance from D. 
The method of measuring by this scale does not altogether get over 
the objection attached to the use of a mental scale, although that 
VOL. XII. 
D 
