ELEMENTS OF SPEECH IN RELATION TO POETRY 45
The difference between these various averages is very slight,
and in every case the variations from it are great. But they are
usually greater for those which have been obtained while the
subjects were in a pleasant state of consciousness than otherwise.
Neutrality and unpleasantness appear to work up to a better
level in the motor consciousness than does pleasurability ; fur¬
thermore, introspectively, there are more varied states of pleas¬
ure than of unpleasantness; if what the subjects gave intro¬
spectively is of any importance, this appears to be empirically
substantiated,—they got after a while to be very reticent about the
introspecting upon the unpleasant combinations, saying only such
things as “very bad,” “I do not care for that at all,” and the like.
But when they got a pleasant combination, they would even wan¬
der into forbidden fields of introspection and bring back material
which had apparently no connection with the subject in hand. Not
all of them, however, but it is quite the fact that the subjects who
found most of these combinations pleasant, show up the most
negatively in these correlations ; at least those who got into the
most effervescent states of mind offer those numerical results
which are the most recalcitrant to satisfactory correlation.
Hints of a Tonal Calculus
We now turn to the graphs for these experiments. Tiren and
Tireen were taken together; and Unrin and Unreen also; these
were drawn in pairs to show the differences obtained from those
experiments which were the nearest alike. We treated in the
same way the graphs for Tira and Thira, and also for Rinaz and
Rinad. The remaining nineteen graphs were arranged in groups
for similar purposes of comparison.
Comparing Tiren and Tireen, it appeared that the short “e” in
Tiren was responsible for the elevation of this graph above the
other. Apparently, also, the “long’ E did not “live up to its
privilege,” for the graphs were of exactly the same length, even
group by group. But the Tiren exceeded the other graph in
height only at the beginning and end of its course. Also, the
effect of the Tireen was steadier than that of Tiren. The latter
showed an average rise toward the fourth iambic and then a