Bayswater.
_Stereoscopic Angles._--When I last addressed you, I fancied I should set
the stereoscopic-angle question at rest. It appears, however, that MR. G.
SHADBOLT is unconvinced, and as I alone (to the best of my knowledge) have
defined and solved the problem in relation to this subject, you will
perhaps allow me to offer a few words in rejoinder to MR. S.'S arguments
which, had that gentleman thought more closely, would not have been
advanced. This is also requisite, because, from their speciousness, they
are likely to mislead such as take what they read for granted. MR. S. says
that when the stereographs are placed at the same distance from the eyes as
the focal length of the lens, that 2-1/4 inches is the best space for the
cameras to be apart; and that were this space increased, the result would
be as though the pictures were taken from models. To this I reply, that the
only correct space for the cameras to be apart is 2-1/2 inches (_i. e._ the
space usually found to be from pupil to pupil of our eyes), and this under
every circumstance; and that any departure from this must produce error. As
to the model-like appearance, I cannot see the reason of {276} it. Next MR.
SHADBOLT says, and rightly, that when the pictures are seen from a less
distance than the focal length of the lens, they appear to be increased in
bulk. But the "obvious remedy" I pronounce to be wrong, as it must produce
error. The remedy is nevertheless obvious, and consists in placing the
stereographs at the same distance from the eyes as the focal length of the
lens. But, if this cannot be done, it were surely better to submit to some
trifling exaggeration than to absolute deformity and error. MR. S. says
also, that as we mainly judge of distance, &c. by the convergence of the
optic axis of our eyes (Query, How do persons with only one eye judge?),
so, in short or medium distances, it were better to let the camera radiate
from its centre to the principal object to be delineated. The result of
this must be error, as the following illustration will show. Let the sitter
(for it is especially recommended in portraits) hold before him,
horizontally, and in parallelism with the picture, a ruler two feet long;
and let planes parallel to the ruler pass through the sitter's ears, eyes,
nose, &c. The consequence would be that the ruler, and all the other planes
parallel to it, would have two vanishing points, and all the features be
erroneously rendered. This, to any one conversant with perspective, should
suffice. But, as all are not acquainted with perspective, perhaps the
following illustration may prove more convincing. Suppose an ass to stand
facing the observer; a boy astride him, with a big drum placed before him.
Now, under the treatment recommended by MR. G. SHADBOLT, both sides of the
ass would be visible; both the boy's legs; and the drum would have two
heads. This would be untrue, absurd, ridiculous, and quite as wonderful as
Mr. Fenton's twelve-feet span view from across the Thames.
Once more, and I shall have done with the present arguments of MR. G.
SHADBOLT. He says that the two pictures should have exactly the same range
of vision. This I deny: for, were it so, there would be no stereoscopic
effect. Let the object be a column: it is evident that a tangent to the
left side of the column from the right eye, could not extend so far to the
left as a tangent to the left side of the column from the left eye, and
_vice versa_. And it is only by this difference in the two pictures (or, in
other words, the range of vision) that our conceptions of solidity are
created. This is not exactly the test to suit the views of MR. SHADBOLT, as
I am quite aware; but I chose it for its simplicity, and because it will
bear demonstration; and my desire has been to elicit truth, and not to
perpetuate error.
In conclusion, I beg to refer MR. G. SHADBOLT to my definition and solution
of the stereoscopic problem--which I then said I _believed_--but which I
now unhesitatingly _assert_ to be correct.