insanity have felt the need of a practical classification of
reactions, and have at the same time encountered the difficulty of
establishing definite criteria for distinguishing the different groups
from one another. It is a comparatively simple matter to make these
distinctions in a general way and even to formulate a more or less
comprehensive theoretical classification, but there still remains much
difficulty in practice. We have made repeated attempts to utilize
various systems of classification which involve free play of personal
equation in their application. Although for us the matter is greatly
simplified by the elimination of all the common reactions with the aid
of the frequency tables, we have nevertheless met with no success. The
distinctions made by either of us have on no occasion fully satisfied,
at the second reading, either the one who made them or the other,
while a comparison of the distinctions made by each of us
independently has shown a disagreement to the extent of 20-35 per
cent.

We sought, therefore, to formulate a classification in which the
various groups should be so defined as to obviate the interference of
personal equation in the work of applying it, hoping thus to achieve
greater accuracy. In this we can lay claim to only partial success;
for, in the first place, having satisfactorily defined a number of
groups, we found it necessary in the end to provide a special group
for unclassified reactions, into which falls more than one-third of
the total number of individual reactions; and, in the second place, in
at least two of our groups the play of personal equation has not been
entirely eliminated, so that there is still a possibility of error to
the extent of five per cent of individual reactions, which means
approximately one per cent of the total number of reactions. We have
found, however, that in spite of these shortcomings the classification