ion, and thereby occasioned a curious
optical illusion, which may perhaps help to shed some light upon the
phenomena now under consideration. For when the sun was high in the sky
and the French window was set at a certain angle, the whole of the
flowers, figures, etc., on my right hand appeared reflected upon the
lawn on the left hand as vividly as if they actually existed in
duplicate. So real was the illusion that for some hours I was under the
impression that a broad yellow gravel path actually stretched across the
lawn on my left. It was only when a little dog ran along the spectral
path and suddenly vanished into thin air that I discovered the illusion.
Nothing could be more complete, more life-like. The real persons who
walked up the gravel to the house walked across the spectral gravel,
apparently in duplicate. Both could be seen at one and the same time. I
instantly thought that they could be photographed, so as to show the
duplication produced by the illusion. Unfortunately, although the
spectral path was distinctly visible through the glass to the eye, no
impression whatever was left on the sensitive plate. My friend writes:--

"I have tried the phantom path, and I am sorry to say it is too phantom
to make any impression on the plate. All that you get is the blaze of
light from the glass window, some very faint trees, and no path at all.
Possibly, with a June sun, it might have been different; but I doubt it,
as one is told never to put the camera facing a window. It is having to
take through the glass window which is fatal."

This set me thinking. It was a simple optical illusion, no doubt,
similar to that which enabled Pepper to produce his ghosts at the
Polytechnic. But what was the agency which enabled me to see the figures
and flowers, and trees and gravel, all transferred, as by the cunning
act of some magician, from the right to the left? Simply a swinging pane
of perfectly transparent glass. To those who have neither studied the
laws of optics nor seen the phenomenon in question, it must seem
impossible that a pellucid window-pane could transfer so faithfully that
which happened at one end of the garden to the other as to cause it to
be mistaken for reality. Yet there was the phenomenon before my eyes.
The dog ran double--the real dog to the right, the spectral dog to the
left--and no one could tell at first sight "t'other from which." Now,
may it not be that this supplies a suggestion as to the cause