66 Haunted Houses and their Phenomena : [February 
nature, and are due to some illusion on the part of the 
observer. The fadds already adverted to — that haunting de- 
creases with the improvement of domestic architecture, with 
the increase of population and of intelligence, with better 
lighting in*doors and out, with the construction of broad 
open roads not overhung by high hedges and trees ; the no 
less decisive faCt that whilst large edifices, public and private, 
are often haunted, cottages and huts never are — all point in 
the same direction, viz., to imperfeCt observation on the part 
of the witnesses. 
We see something in dim light. Then, if our fancy or 
our fears have been excited by loneliness, by the size, the 
antiquity, or the traditions of the place where we happen to 
be, we add, without the least intent to deceive, strange de- 
tails to the image impressed upon our retina. 
Illusions similar to those experienced by Nicolai, and due 
not to any outward objeCt, but to a morbid condition of the 
brain, are perhaps more common than is suspected. As far 
as I am aware no one, on seeing a supposed ghost, has ever 
applied to it the ingenious test suggested by the late Sir D. 
Brewster. He pointed out that if we press the eye-ball 
slightly with the finger any real objeCt placed before us will 
appear double, whilst any phantom having no substantive 
existence will remain single. 
That by far the greater number of supposed cases of 
haunted houses, or rather the phenomena on the faith of 
which such houses are pronounced haunted, must be regarded 
as subjective is certain ; that in a very large portion of the 
remainder the occurrences on record may be traced to natural 
intelligible causes is equally decided ; but there is a balance 
upon which it is unphilosophical to pronounce. Let us re- 
member that if an avowed juggler reproduces before our eyes 
some surprising phenomenon which we have witnessed, this 
is no proof that the original phenomenon was due to jugglery. 
A skilful mechanician may make an automaton eagle, duck, 
or fly. Is this evidence that eagles, ducks, and flies are 
automata ? An adroit coiner may palm upon you a piece of 
base metal as current coin of the realm. Do you, on disco- 
vering the cheat, pronounce all gold spurious ? 
I may mention an instance of a haunted house related to 
me by a scientific friend, a gentleman chiefly engaged with 
the study of physics and chemistry, and very far from credu- 
lous. When a youth his family resided for some time in 
Bretagne, and had taken on lease a house at somelittle dis- 
tance from a small town. Like many French country seats 
this house faced on one side— the south in this case — on a 
