21 8 Dr. Herschel/s Experiments for investigating 
the alternate colours of the two sets that will then be seen by 
the rays j, 2, 3, and 1, 2, 4, 5. Remove in the next place 
the stop from the looking-glass, and bring the second shadow 
of the penknife over the primary set, and there will then 
only remain the two sets of rings formed by incident rays 
which come from 6, and which have never passed through the 
tipper surface of the lens. Now, as both sets of rings in this 
case are completely formed by rays transmitted upwards 
from the coated part of the looking-glass without passing 
through the first surface of the incumbent lens, the proof 
that the modifying power of that surface is not required to 
the formation of the rings is established. 
It can hardly be supposed that the first surface of the lens 
should have any concern in the formation of the rings when 
the rays are reflected from the looking-glass towards the 
eye ; but the same experiment, that has proved that this 
surface was not required to be used with incident rays, will 
show that we may do without it when they are on their 
return. We need only invert the fractured lens, as in figure 
12, when either the rays 1, 2, 4, 5, or 6, 7, 2, 4, 5, will 
convey the image of the rings after their formation to the eye 
without passing through any part of the lens. 
XXV. Of the Action of the second Surface. 
As rings are formed when two glasses are laid upon each 
other, it is but reasonable to expect that the two surfaces at 
least which are placed together should have an immediate 
effect upon them, and so much the more, as it has been ascer- 
tained that the first surface assists only by permitting light to 
pass into the body of the glass. Some of the experiments 
