24 
* . Transactions of the Society. 
either by contact or by projection, so long as the film faces tbe lens, 
the image will be similar to No. 1.) 
No. 5. By inverting the drawing the original is represented. 
No. 6. The original is not reproduced whichever way the paper is 
turned. In order to obtain an exact copy of the original it is necessary 
to draw the object on tracing paper and view it from the wrong side. 
No. 7. This case is not so clear ; the mirror erects the inverted 
image No. 2, but its transposition is due to the fact of its not being 
viewed as a transparency. 
Addendum ( read 19 th Dec. 1894). 
The discussion on my paper opened out some interesting points with 
regard to drawing with the Microscope and the distortion of the image. 
It was suggested that a direct drawing made by placing ruled squares 
on the diaphragm of the eye-piece, and using paper ruled in squares, 
was more accurate than one made by a camera lucida, for this reason, 
that unless all points of the paper were at an equal distance from the 
camera lucida the magnification would not be uniform throughout the 
image. In other words, that unless a camera lucida drawing is made 
in a hollow spherical shell, whose radius is equal to that of the distance 
of the paper from the camera lucida, distortion will be present. 
This is quite true, but is the direct method any better ? for it was 
pointed out many years ago by Coddington, that owing to the “ over 
refraction of the oblique pencils ” a 
real image had barrel distortion, 
fig. 2, and a virtual pincushion dis- 
tortion, fig. 3. The real image 
formed by the objective has therefore 
barrel distortion, fig. 2, which is 
made rectilinear by the virtual or 
pincushion distortion of the eye- 
piece. If therefore we place ruled 
squares on the diaphragm of the 
eye-piece they will suffer pincushion distortion by reason of the virtual 
image formed by the eye-lens. Hence if we place ruled squares on 
the stage of the Microscope, and view their image through a properly 
constructed Microscope, we ought to obtain a perfectly rectilinear 
image, because the barrel distortion will have been compensated by the 
pincushion distortion of the eye-piece. But if we make a drawing of 
this rectilinear figure by means of the pincushion distorted squares at 
the diaphragm of the eye-piece, the drawing so made will obviously be 
barrel distorted. Suppose we place an objective on the nose-piece, 
which will magnify the squares on the stage so that they appear the 
same size as those at the diaphragm of the eye-piece, they will coin- 
cide with one another in the centre of the field, but at the margin, 
where the pincushion distortion more largely exists, the lines of the 
Fig. 2. 
□ZED 
LL. 
Fig. 3. 
