448 
EXPLANATION OF THE CYLINDRIC PRINCIPLE, 
22 July 
when I told him that it was in order to show his town to my friends 
at home in my own country, and to let them see how extensive it was. 
He took much interest in my operations, and conducted me to a 
spot whence I could have the best prospect. 
His choice was just, and I made my drawing from the very 
point of view to which he brought me ; but the numerous clusters 
of dwellings spread so far both to the right and to the left, that the 
laws of perspective would not allow me to include in my sketch 
more than a third of the town, without having recourse to the prin- 
ciple of a cylindric tuedium. * But as there was no variation in the 
* The method of dramng in perspective on the principle of an imaginary cylindric 
medium not having hitherto, I beHeve, been noticed by any writer on this branch of optics, 
it becomes necessary to make the above allusion to it intelligible by a brief explanation. — 
The usual method supposes this medium to be a transparent plane through vi^hich 
the objects are beheld ; the ' point of sight,' to be that point in the plane, where a line 
from the eye, or visual ray, would strike it perpendicularly ; the ' point of distance,' to be 
a point in the plane, at a distance from the point of sight, equal to that between the eye and 
the medium ; and ' accidental points' to be those to which all lines not actually parallel to 
the medium, nor perpendicular to it, appear perspectively to converge. This, at least, is the 
method usually taught, and is that which I acquired under the instruction of that excellent 
artist Mr. Nattes, whose works prove that he was truly a master in the art of perspective 
drawing. In putting this method into practice, the view seen through the medium, or 
rather the medium itself on which the objects are imagined as depicted, becomes the 
picture ; which picture, to be strictly and optically correct, must represent these objects 
larger in proportion as their places on the medium may be farther from the point of sight ; 
an enlargement similar to that which is requii'ed in the projection of extensive maps. 
Until the extent of landscape exceed an angle of about fifty degrees, this enlarge- 
ment will be inconsiderable in drawings of moderate size ; but when it is much beyond 
that, the picture, if the correct principles of perspective be strictly adhered to, must be- 
come an anamorphosis, and will not admit of being viewed any other position, than 
that in which the eye of the spectator is exactly opposite to the ' point of sight,' and at 
the same distance from it, as the ' point of distance.' This is, however, the position in 
which all perspectives ought to be viewed : notwithstanding a practice, too often seen in 
galleries, of hanging such pictures so that the visual ray cannot fall perpendicularly upon 
its proper point, nor even upon any part of the painting. 
But by adopting that principle which supposes this medium to be the superficial con- 
cavity of a perpendicular cylinder, the eye being placed at any point in the axis, a method 
is found by which a landscape may be extended to any number of degrees, and by which 
every object on the same horizontal line may be delineated in the same proportion ; which 
may be demonstrated geometrically by showing that all the visual rays from the axis of a 
cylinder, and falling on the same horizontal line, are equal ; while all those which fall upon 
a nlane, increase in length as they spread farther from the point of sight. In practice. 
