128 
Transactions of the Society . 
Table II. 
Line. 
Wave-length 
in inches. 
Wave-length 
in 
tenth-metres. 
A 
* 
Reciprocal of 
Wave-length 
in 
tenth -metres 
X 1000. 
L 
X 
— 
A' 
1 
33086 
7677-0 
•1302592 
B 
1 
36971 
6870-19 
•1455564 
C 
1 
38702 
6563-05 
•1523681 
1) 
1 
43102 
5893-0 
•1696929 
Middle of L> lines. 
1 
43122 
5890-18 
•1697741 
1 
45000 
1 
48193 
5644-44 
•1771655 
Point of maximum 
visual brightness. 
E 
5270-5 
•1897353 
F 
1 
52247 
4861-5 
•2056978 
G' 
1 
58512 
4341 '0 
•2303617 
1 
64002 
3968-62 
•2519767 
fore be exceedingly long, and a lens for practical work would require 
to have curves of excessively short radii. In order to obtain radii 
suitable for practical constructions, it is necessary to choose glasses 
having a dispersive ratio of 1 • 25 and upwards, unless the lens be 
required of very long, or even of infinite, focus for purposes of correc- 
tion. 
Now, with regard to the small figures printed below the others, 
they are for the purpose of estimating the amount of secondary spec- 
trum left outstanding in the combination. 
The presence of a secondary spectrum is owing to irrationality 
in the dispersion. To illustrate this, let us suppose that we have a 
flint, the dispersion of which between B and E is precisely double 
