Briefly summing up we see that when a point source is employed the slit of the spectroscope is illuminated by light coming from one point only in the wedge, and consequently clear distinct bands will be obtained in the spectrum irrespective of which end of the wedge is towards the source, and irrespective of the angle of incidence or of the distance between the wedge and spectroscope.

With a large source, however, the slit of the spectroscope receives rays coming from many different points in the wedge. The path difference between the interfering components of each of these rays is in general different. Therefore each ray will produce a complete set of interference bands in the spectrum. These sets of bands are relatively displaced, the superposition of all causing uniform illumination.

It has been shown, however, that there is one special case in which the path difference for the interfering components of all rays is the same. In this case, each ray still produces, a complete set of bands in the spectrum but these sets are no longer relatively displaced; bright bands are superimposed on bright bands and dark bands coincide with dark bands. There is thus both an increase of intensity and definition.

It has been shown that the conditions necessary for this special case are:

(1) The thick end of the wedge is towards the source.

(2) The slit of the spectroscope is placed at a distance of

from the wedge, the distance being measured along the central ray entering the slit of the spectroscope.

The case of an air wedge has been discussed, and a possible advantage to be gained in using such for spectroscopic purposes in place of parallel plates has been mentioned.