readiness to break away from a train
of associations as soon as the "point of diminishing returns" has been
reached. This doubtless explains why adults sometimes make such a
surprisingly poor showing in the test. They have less "intellectual
nonchalance" than children, are less willing to subordinate such
considerations as completeness and logical connection to the demands of
speed. Knollin's unemployed men of 12- to 13-year intelligence succeeded
no better than school children of the 10-year level.

We do not believe, however, that this fault is serious enough to warrant
the elimination of the test. The fact is that in a large majority of
cases the score which it yields agrees fairly closely with the result of
the scale as a whole. Subjects more than a year or two below the mental
age of 10 years seldom succeed. Those more than a year or two above the
10-year level seldom fail.

There is another reason why the test should be retained, it often has
significance beyond that which appears in the mere number of words
given. The naming of unusual and abstract words is an instance of this.
An unusually large number of repetitions has symptomatic significance
in the other direction. It indicates a tendency to mental stereotypy, so
frequently encountered in testing the feeble-minded. The proportion of
repetitions made by normal children of the 10- or 11-year intelligence
level rarely exceeds 2 or 3 per cent of the total number of words named;
those of older retarded children of the same level occasionally reach
6 or 8 per cent.

It is conceivable, of course, that a more satisfactory test of this
general nature could be devised; such, for example, as having the
subject name all the words he can of a given class (four-footed animals,
things to eat, articles of household furniture, trees, birds, etc.). The
main objection to this form of the test is that the performance would in
all probability be more influenced by environment and formal instruction
than is the case with the test of naming sixty words.

One other matter remains to be mentioned; namely, the relative number of
words named in the half-minute periods. As would be expected, the rate
of naming words decreases as the test proceeds. In the case of the
10-year-olds, we find the average number of words for the six successive
half-minutes to be as follows:--

18, 121/2, 101/2, 9, 81/2, 7.

Some subjects maintain an almost constant rate throughout the test,
other