6 
At threshold, Taylor would predict an annulus width of 1.25 
minutes at the adaptation level used in these investigations, but 
the width could be expected to be somewhat less in these tests since 
the targets were well above threshold. 
The steep linear portion at the left of the curves in figure 3 
suggests that this annulus is significant in pointing, and has a width 
of approximately 1 minute under the conditions described. Although 
the subjective annulus was presumably not present with the smaller 
measuring marks, fixation would cause the judgement to be made on the 
sensitive central fovea. For the larger measuring marks, and the 
associated subjective annuli, the image excursion caused by nystagmus 
would bring all parts of the annulus on to the central fovea in turn, 
providing the information for the epicritic judgement. 
Figure 3 suggests a zone of constancy approximately 4 minutes 
of arc in diameter, but this seems to be a function of the annulus 
between mark and target rather than the overall visual angle of the 
target. An annulus width of 0.15 minutes gives a pointing error of 
the order of 1.2 seconds irrespective of the overall visual angle up 
to a diameter of at least 45 minutes. 
It will be noted that the pointing error Sp is a function S x 
and Sy. The pointing error in the X and Y directions may be esti 
mated from the curves as For annuli of width 0.15 minutes 
of arc, this amounts to approximately 0.8 seconds. This could pre 
sumably be expected in theodolite pointings, or in any optical 
settings under the same conditions. 
As the width of the annulus increases above approximately 4 
minutes of arc the curves rise and approach the corresponding Weber's 
law curves obtained earlier with the stereocomparator (10). 
III. CONCLUSIONS 
These investigations were restricted to foveal imagery of high- 
contrast circular targets and measuring marks subtending visual angles 
of 2-45 minutes of arc. This range should include the sizes likely 
to be used for artificial pass-points for visual pointing. 
The thresholds obtained were in good agreement with generally 
accepted vernier acuity thresholds, and appeared to justify the de 
sign of the special apparatus. 
The results supported the concept of a retinal zone approxi 
mately 4 minutes of arc in diameter over which acuity is constant, 
but further suggested that this may be dependent on the type of task 
involved. The maximum information content for the pointing appeared