from it to disturb the
Experiment, and viewing the Paper through a third Prism held parallel
to it, I saw that half of it which was illuminated by the violet Light
to be divided from the other half by a greater Refraction, especially
when I went a good way off from the Paper. For when I viewed it too near
at hand, the two halfs of the Paper did not appear fully divided from
one another, but seemed contiguous at one of their Angles like the
painted Paper in the first Experiment. Which also happened when the
Paper was too broad.

[Illustration: FIG. 19.]

Sometimes instead of the Paper I used a white Thred, and this appeared
through the Prism divided into two parallel Threds as is represented in
the nineteenth Figure, where DG denotes the Thred illuminated with
violet Light from D to E and with red Light from F to G, and _defg_ are
the parts of the Thred seen by Refraction. If one half of the Thred be
constantly illuminated with red, and the other half be illuminated with
all the Colours successively, (which may be done by causing one of the
Prisms to be turned about its Axis whilst the other remains unmoved)
this other half in viewing the Thred through the Prism, will appear in
a continual right Line with the first half when illuminated with red,
and begin to be a little divided from it when illuminated with Orange,
and remove farther from it when illuminated with yellow, and still
farther when with green, and farther when with blue, and go yet farther
off when illuminated with Indigo, and farthest when with deep violet.
Which plainly shews, that the Lights of several Colours are more and
more refrangible one than another, in this Order of their Colours, red,
orange, yellow, green, blue, indigo, deep violet; and so proves as well
the first Proposition as the second.

I caused also the coloured Spectrums PT [in _Fig._ 17.] and MN made in a
dark Chamber by the Refractions of two Prisms to lie in a Right Line end
to end, as was described above in the fifth Experiment, and viewing them
through a third Prism held parallel to their Length, they appeared no
longer in a Right Line, but became broken from one another, as they are
represented at _pt_ and _mn_, the violet end _m_ of the Spectrum _mn_
being by a greater Refraction translated farther from its former Place
MT than the red end _t_ of the other Spectrum _pt_.

I farther caused those two Spectrums PT [in _Fig._ 20.] and MN to become
co-incident in an inverted Order