hanical memory.

This test was given a place in year VII of Binet's 1908 scale, the coins
used being the 1-sou, 2-sous, 10-sous, and 5-franc pieces. It was
omitted from the Binet 1911 revision and also from that of Goddard.
Kuhlmann retains it in year VII. Others, however, have required all four
coins to be correctly named, and when this standard is used the test is
difficult enough for year VII. Germany has six coins up to and including
the 1-mark piece, all of which could be named by 76 per cent of
Bobertag's 7-year-olds. With the coins and the standard of scoring used
in the Stanford revision the test belongs well in year VI.

VI, 6. REPEATING SIXTEEN TO EIGHTEEN SYLLABLES

The sentences are:--

(a) "_We are having a fine time. We found a little mouse in the
trap._"
(b) "_Walter had a fine time on his vacation. He went fishing
every day._"
(c) "_We will go out for a long walk. Please give me my pretty
straw hat._"

PROCEDURE. The instructions should be given as follows: "_Now, listen. I
am going to say something and after I am through I want you to say it
over just like I do. Understand? Listen carefully and be sure to say
exactly what I say._" Then read the first sentence rather slowly, in a
distinct voice, and with expression. If the response is not too bad,
praise the child's efforts. Then proceed with the second and third
sentences, prefacing each with an exhortation to "say exactly what I
say."

In this year and in the memory-for-sentences test of later years it is
not permissible to re-read even the first sentence. The only reason for
allowing a repetition of one of the sentences in the earlier test of
this kind was to overcome the child's timidity. With children of 6 years
or upward we seldom encounter the timidity which sometimes makes it so
hard to secure responses in some of the tests of the earlier years.

SCORING. The test is passed _if at least one sentence out of three is
repeated without error, or if two are repeated with not more than one
error each_. A single omission, insertion, or transposition counts as an
error. Faults of pronunciation are of course overlooked. It is not
sufficient that the thought be reproduced intact; the exact language
must be repeated. The responses should be recorded _verbatim_. This is
easily done if record blanks used for scoring have the sentences printed
in full.

REMARKS. In this test and in later tests of memory for sentences, it is