90 J. LeConte— Phenomena of Binocular Vision.” 
torsion of the eye, but the inclination of the horizontal image 
on the a lines of the wall does not give the true tor- 
sion of the 
here are many other ways of testing the truth of this last 
proposition and the falsity of the reverse statement of Helm- 
holtz. If we make, as before, a vertical image, and instead 
of turning the eyes upward and to the right, turn the body 
to the right and the face upward and cast “the image on the 
extreme Tight and upper portion of the wall; the vertical 
image will “be projected vertically on the wall, but a hon- 
zontal image cast to the same pl: ice, in the same way, will 
be inclined in the same way as in Helmholtz’s diagram, but 
at much greater angle. In this case, the eyes are in the primary 
position, and therefore there is no rotation at all, the inclina- 
tion of the horizontal image is the result of projection alone. 
Without any attempt at mathematical accuracy, the dia- 
gram, figure 2, shows the manner in which spheric: ul coor: 
dinates would project on a plane perpendicular wall. The 
crosses in the corners show how a rectangular cross image 
~ 
wunkd be distorted by projection alone. Now by careful plot- 
ting, I have rane that at a point 40° upward or downward, 
40° to one side right or left, the inclination 
of the hyperbolic curves with the true horizontals 
of the wall is about 20°—which makes the angles 
of the projected cross 70° and 110°. The rotation 
| of such a cross 15°, would give exactly the results 
Hi obtained by experiment. In figure 8, the heavy 
sross shows the position of the image when dis- 
torted by pen only, -~ the lighter lines ‘the same as 
rotated 15° to the right. As the result of this rotation, the 
vertical line is inclined 15° to the right, while the horizontal line 
is inclined only 5° ¢éo the left, as we found by experiment. 
3. 
/ 
/ 
