Dr Searle, Experiments with a prism of small angle 163 
isd (4D — D, 2 
we find OP oe ee Stee eae (12) 
Subtracting (10) from (9), we find D—D,=(p,—1)7, and thus 
peace PCAs ie A eae: (13) 
fa — 1 
§ 6. Experimental details. The measurements are made by 
the goniometer described in § 2. The collimator AB (Fig. 9) has 
a pair of cross wires at A. ‘These wires are fixed for their pro- 
tection to the inner end of a short tube which slides im the main 
Fig. 9. 
tube of the collimator, and the wires are adjusted to lie in the 
focal plane of the achromatic lens B (of 15 to 20 em. focus). The 
tank C has sides of plane glass, but it is not necessary that the 
tivo sides should be accurately parallel. The collimator is illumi- 
nated by a sodium flame F. If the goniometer D# is properly 
placed, an image of the vertical wire at A will coincide with the 
vertical wire of the goniometer at #. he tank contains the 
liquid of refractive index sy. 
The prism is first placed between the collimator and the 
goniometer, the tank bemg removed, and a reading of the gonio- 
meter is taken. The prism is then turned through 180° about a 
vertical axis and a second reading is taken. The difference 
between the two readings corresponds to 2D. 
The tank is then put into place and is filled with water, and 
the prism is placed in the tank, as indicated in Fig. 9. The 
goniometer reading is taken and then, without disturbing the 
tank, the prism is turned through 180° and a second reading is 
taken, the difference corresponding to 2D,. This reversal of the 
prism eliminates any error due to want of parallelism between 
the sides of the tank. 
If x and 2, be the changes of goniometer reading corresponding 
to D and to D,, the angles will be so small with a suitable prism 
that the corrections given by the Table in § 2 will be negligible, 
and thus we may write 
D=a/2l, D,=2,/21, 
