267 
1912-13.] The Scattering of Light. 
a straight line across the card, the deviations being due to errors of setting. 
The necessary work of averaging is done without any measurement by a 
simple drawing instrument I designed for the purpose. This method of 
working enables both settings and reductions to be most expeditiously 
carried out. To save time in adjusting the telescope of the spectrophoto- 
meter to any required spectral region, a helium tube is mounted on a 
rotatable arm, so that it can be swung round, its light directed into the 
collimator, and the telescope set on the required line, without the necessity 
for reading a scale or even turning up the lights in the room. The angle 
of emission from the ground-glass plate of the light which is passing into 
the collimator can also be regulated quite simply in the dark. The plate 
is fixed to a levelling table standing on the end of the long arm A which 
is pivoted under G 2 so as to be movable round the pivot in a horizontal 
plane. A catch C can be made to engage with any one of the studs 
shown, whose positions are carefully adjusted so as to lock the arm at will 
at the angles 90°, 85°, 80°, etc., with R. On A stands a lantern case 
completely enclosing the Nernst lamp N p furnished with an achromatic 
condensing lens L p from which a parallel beam of light falls upon the ground 
glass G p all the Nernst filament except a small portion at the centre about 
a quarter of an inch long having been screened off. It will be understood 
that in consequence of the above arrangements only those rays of the 
parallel beam incident upon the plate which are scattered as a parallel 
beam at one particular angle enter the collimator slit, and that this angle 
depends upon the position of the arm A. The various adjustments which 
have to be made to ensure that the beam of light falls perpendicularly on 
the surface of the plate, that the mirror M is in the proper position, etc., 
etc., are carried out by means of the usual slow-motion-screw fittings, and 
by the help of a small telescope T of special construction, which remains 
permanently in the position shown, as a means of subsequent verification 
from time to time. For the purpose of enabling experiments to be made 
in which the incident light falls obliquely on the glass plate, it has been 
arranged that the table to which G x is attached can be supported inde- 
pendently of A, so that G x can be turned to any desired angle with the 
latter. Care has been taken to design all the fittings for the levelling 
table, the arm A, and the catch C on “ geometrical ” principles, A for example 
standing on two rollers at the extremities of the crossbar and having at 
the other end as pivot a ball resting in a conical hollow. The result of 
this design is that no looseness has come about through wear, and yet both 
the levelling table and the arm with its lantern can at any time simply be 
lifted up and carried away for any purpose, and as simply replaced, with 
