110 PROGRESS IN MICROSCOPY 



the ground E' is adjustable. In Fig. 3.16, the ordinary and extra- 

 ordinary irnages are both duplicated axially only although, in some 

 instruments such axial duplication is concomitant with lateral some 

 shift that does not, however, affect the principle of the process. 



7. SMITH'S FULL IMAGE DUPLICATION INTERFERENCE MICROSCOPE 



This type of microscope is based on the use of WoUaston prisms. 

 It is well known that the latter consist of two quartz prisms the axes 

 of which are crossed (Fig. 3.17). One axis lies in the plane of said 



Fig. 3.17. Wollaston prism. 



figure, the other is perpendicular to it. An incident light-ray, normal 

 at the ingoing face, is split, at M, in two rays: the ordinary ray O and 

 the extraordinary ray E. Phenomena occur as in Figs. 3.9 and 3.12 

 but the linear duplication is replaced by an angular one. If the point 

 M lies in the area where both prisms have the same thickness, the 

 outgoing rays are in phase. If the incident ray impinges the prism 

 at M' the outgoing rays O and E are no longer in phase. Therefore, 

 the phase difference between the rays O and E may be varied according 

 to the Wollaston area used. Observing a Wollaston prism between 

 two polarizers there is a whole series of fringes that are parallel, 

 equidistant and straight in relation to the Wollaston prism edges. 

 In white light and between crossed polarizers there is a dark fringe 

 edged on both sides by colour fringes. The colour sequence is sub- 

 stantially that in Newton's scale of colours. The dark fringe is at M 

 in the area where both prisms have the same thickness. Now let us 

 image a narrow source at M on the dark fringe (Fig. 3.18) and let 

 us assume the beam's aperture angle a to be narrow. The rays O are 

 in phase with the rays E (such would not be the case were the angle m 

 to exceed a few degrees). Since both polarizers £?i and £?2 ^ire crossed, 

 no trace of the beam on the screen T remains. In Fig. 3.18, duplication 

 of the emerging beam is not shown: it is assumed that such duplication 

 is minimal in relation to the beam width. Let us move the Wollaston 

 prism at right angles to the mean ray of the incident beam, e.g. in 



