334 W. B. Rogers on Binocular Vision. 
responding ends of the lines, and that thé'tsvo resultant points or 
short Vertical lines thus formed are produéed in the same way as 
the two resultants of; unequally distant verticals treated of under 
a preceding head. The intervening points of the horizontal lines 
being all alike are not, as in the case of the extremities, marke 
out in corresponding pairs, and therefore may be combined binoc- 
ularly in various ways, according to the rate of variation of the 
axial convergence in the interval between the terminal combina- 
tions. On a close examination of the resultant, I find accord- 
_ingly, that while the actual amount of its relief remains constant, 
the intermediate parts of the line do not always appear in a 
straight direction between the near and remote end. ‘Thus sup- 
posing the combination to be made in front of the paper MN (fig. 
43) and that C and D represent the points at which the corres- 
43- 
ponding ends Aa and Bb appear combined, I observe some- 
times that the resultant loses the clear and satisfactory character 
- : 
evanescent appearances are I think produced by variable con- 
verging movements in the same interval. 
rom these experiments, we are I think entitled to conclude 
that unequal horizontal lines not greatly differing in length may 
be combined so as to form a perspective resultant, the right and 
left ends of which are at unequal distances from the observer. 
This result, which appears to have escaped the attention of pre- 
ceding experimenters, is of importance from the light thrown by 
it on the binocular union of unequal figures. In this connexion 
it will be referred to under a future head, but in the mean time 
the following example of its application to this question may 
properly conclude the present division of our subject. 
