SENSE ORGANS OF BIRDS—-PUMPHREY 
insect is presumably viewed binocu- 
ularly, and the action is controlled by 
a distance appreciation based on the 
amount of accommodation required to 
bring the image of the insect into focus; 
but it is likely that the parallax esti- 
mate is necessary before the head can 
safely be brought into position for a 
peck, and that the successive scrutinies 
are not due to uncertainty about the 
edibility of the food.* 
As has already been indicated, mon- 
ocularity is considerably modified in 
those diurnal birds which pursue mov- 
ing prey, but even in these binocular 
fixation appears only to occur as a 
response to a moving object which is 
close athand. Remote moving objects 
are scrutinized monocularly, perhaps 
because their movement is more per- 
ceptible when the image falls on the 
central fovea. Prey appear to be sought 
monocularly, and only in the pursuit 
is binocular fixation brought into play. 
But although we may suspect that 
avian vision is fundamentally based on 
monocularity, operations in which the 
beak is involyed—pecking, preening, 
nest building, and so forth—have en- 
sured that there was always a use for 
the anterior binocular field, and that 
the potentiality of a more highly 
developed binocularity was always la- 
tent. When it emerges, as, judging by 
the occurrence of a temporal fovea, it 
has done independently in many re- 
motely related birds, it does so in strict 
conformity with the requirements of 
the bird’s habits, and would, presuma- 
bly, subside again if the abandonment 
of a raptorial career rendered it 
unnecessary. 
VISUAL ACUITY OF BIRDS 
It is known that the foveal acuity of 
man is very close to the maximum 
physically possible. The imperfection 
of the image formed by the lens system 
4I am indebted to R. E. Moreau for the 
information that an egret, feeding on insects 
on the ground, also appears to make several 
parallax estimates before pecking. 
315 
and the coarseness of grain of the 
retina individually impose limits on 
what the eye can resolve, but it is 
reasonable to expect these limits to be 
the same, for otherwise either the 
retina or the dioptric system would be 
better than it need be, contrary to the 
law of parsimony that natural selec- 
tion rigorously enforces. It is found in 
fact that two self-luminous points 
such as stars can in the best conditions 
be seen as separate when their angular 
separation is more than (but less than 
twice) the angular separation of two 
adjacent cones in the fovea, which is 
about 30 seconds of arc. Moreover, 
human acuity is somewhat improved 
by the use of monochromatic illumi- 
nation and is diminished by a decrease 
in pupil width. The former elimi- 
nates chromatic aberration and the 
latter increases the extent of the 
diffraction pattern which forms the 
image of a point in the field, so we 
may conclude that both chromatic 
aberration and diffraction are limiting 
factors in the human eye. We find 
however, that in the central fovea of 
buzzards the effective angular separa- 
tion of the cones is only about one- 
third that of the cones in the human 
fovea. Whether the bird fovea is con- 
cerned principally with discrimination 
of movement or principally with dis- 
crimination of detail, in either event 
the high cone-density must be sup- 
posed to be of functional value. And 
its advantages can only be realized if 
the dioptric system is improved pro- 
portionately, that is to say, the aper- 
ture—the ratio of pupil diameter to 
focal length—must be about three 
times the aperture of the human eye 
and, in spite of this, there must be a 
proportionate decrease in the effect of 
chromatic aberration. 
The shape of the bird eye makes the 
necessary pupil width quite possible, 
and though diurnal birds show pupillo- 
motor reflexes (contractions of the pupil 
in response to sudden increases in il- 
lumination), the contractions are not 
maintained as they are in man, and the 
