OF NITRO-CELLULOSE AND THE LAW OF ITS OPTICAL BEHAVIOUR. 
145 
observed directly by eye instead of projecting on to a screen. The general arrange¬ 
ment of the apparatus is shown in the accompanying fig. 4, in which a plane polarised 
beam of white light from a Nicol’s prism A is transmitted through the tension 
specimen B, to which an extensometer C is secured, and is then focussed by a lens D 
on a horizontal slit in order that the light passing through the comparison beams 
B.mt.due to weight of lever, 
pendant, etc.= z-5 2 lh.“inches 
shall be at the same level throughout. This thin pencil of light is again brought 
to parallelism before passing through the compound beam F and analyser G, and is 
finally focussed on a ruled glass slide H provided with an eye-piece J. The weight 
of the extension beams and hangers causes a bending moment in the beams which 
has been allowed for in all calculations of stress. In order to compare the different 
specimens one with another, an “ equivalent stress ” in each specimen is calculated, 
that is, such a stress as would produce the same relative retardation in a piece of 
nitro-cellulose of the same material as the standard beam, but of the thickness of the 
specimen under observation. Thus if the thickness of specimen is t and the stress in 
