20 
Frank Angell. 
to perceive a difference. In accordance with this we find that 
reaction-times of A1 for »like« are slower for all time inter¬ 
vals, both with and without D’n. When a series of comparisons is 
made up of stimuli, differing in part hut a little, and in part not 
at all, from the norm stimulus, the judgments of »like« may he 
attended by a conviction of likeness which is very often due to a 
feeling, mood or tension sensation, or even to accidental circum¬ 
stance. The judgment »like« which A1 delivered with the strongest 
feeling of conviction, came from a faint pink tinge, on both norm 
and comparison, resulting from a trace of blue in the page of the 
note-book. If then mental processes, not necessarily integral parts 
of the visual image of the disc, hut present both at the exposure of 
the norm and the comparison, may determine judgments of »like«, 
then we should expect the greatest frequency of these judgments 
where such common factors most frequently occur, viz. in the shorter 
time intervals. Turning to the averages of the undistracted intervals 
of table III, we find that the average number of »like« judgments 
of the 5 sec. intervals for A1 is 21,2 % ; for the other three inter¬ 
vals 11,4 %. For Bt the corresponding figures are: 5 sec. —19,4 % ; 
other intervals — 10,4 %. 
This last fact helps to explain the greater proportion of right 
judgments of »like« in the shorter intervals. Granting, as is highly 
probable, that more direct and accurate acts of visual comparison 
take place in the shorter intervals, we have also the additional factor 
of the relatively greater number of these judgments in the shorter 
intervals. 
It is obvious that the tendency towards incorrect judgments will 
be greater for like than for unlike values of N and V for the longer 
time intervals. For N and V in these judgments are medium shades 
and less apt to call out free judgments than the extreme brightnesses 
used as comparisons. But the effect on judgments depending on 
contiguous association would probably be still more pronounced: for 
in the first place, a medium shade of gray might be classed at one 
time as light and another time as dark, hut neither with any great 
degree of conviction so that in the time order N—V especially the 
associated member whether verbal or otherwise, might easily he con¬ 
fused or lost. This is shown clearly in the case of Bt who makes