1415 
or less some of the first set as to form, or colour or both. Next the 
observer was required to image those original figures of which a 
fragment would be shown him, and to transter his mental repro- 
ductions to paper while reporting the various details of form and 
colour as faithfully as could be. Our object in repeating the im- 
pression of some of the original stimuli was to see whether and 
how far the recognition was acted upon by them, and thus to secure 
a touchstone by which to test the influence of the mental repro- 
duetions upon the recognition. : 
Both influences were the object of an investigation at a third sitting, 
which again took place 24 hrs. later. The primary stimuli were 
presented again to the observer, who was then required to say 
whether or not he recognized them. The time needed for reeogniuon 
was registered by means of a chronometer. 
In order to ascertain the extent to which the recognition had been 
facilitated by the impression of the fragments of the original stimuli, 
applied at the second sitting, the two observers were subjected to a 
set of control-tests. Here also our material was uniform and meaning- 
less. After familiarizing themselves with twelve figures of the kind, 
in the way indicated heretofore, the experimenter again selected five, 
of which fragments were shown at a second sitting, the subject 
being again called on to avoid, as much as possible, any reproduction 
of the complete figure. Then again a third sitting followed for the 
recognition of the original stimuli. To throw the observer into 
confusion, the intervals between the experiments proper, as well as 
those of the control-tests were filled up with “Vexirversuche”’, so that 
the observer could not tell beforehand whether the stimulus presented 
for recognition, was or was not one of the familiar series. 
Each observer was subjected to 180 experiments, 60 of which 
served for comparison. 
The drawings made during the second sitting, fell into three groups 
upon the basis of their likeness to the primary figures, viz. accurate, 
fairly accurate and inaccurate drawings. A drawing was considered 
accurate, when it was in every detail a faithful copy of the original 
figure; fairly accurate when it differed from the latter in certain 
details as to position and extensions of the parts. Such drawings as 
showed an essential departure from the principal features of the primary 
stimulus were called inaccurate. Our criterion is, it is true, open to 
objection, but in our judgment it js the only one at our disposal. 
In the following table the drawings of our subjects are grouped 
under the 3 headings mentioned above. The colour-sensation has 
been subtracted, since as to colours the reproductive tendencies are 
91* 
