REMARKS. After the three copies have been made say: "_Which one do you like best?_" In this way we get an idea of the subject's power of auto-criticism, a trait in which the mentally retarded are nearly always behind normal children of their own age. Normal children, when young, reveal the same weakness to a certain extent. It is especially significant when the subject shows complete satisfaction with a very poor performance. Observe whether the child makes each part with careful effort, looking at the model from time to time, or whether the strokes are made in a haphazard way with only an initial glance at the original. The latter procedure is quite common with young or retarded subjects. Curiously enough, the first trial is more successful than either of the others, due perhaps to a waning of effort and attention. Note that pencil is used instead of pen and that only one success is necessary. Binet gives only one trial and requires pen. Goddard allows pencil, but permits only one trial. Kuhlmann requires pen and passes the child only when two trials out of three are successful. But these authors locate the test at 5 years. Our results show that nearly three fourths of 4-year-olds succeed with pencil in one out of three trials if the scoring is liberal. It makes a great deal of difference whether pen or pencil is used, and whether two successes are required or only one. No better illustration could be given of the fact that without thoroughgoing standardization of procedure and scoring the best mental test may be misleading as to the degree of intelligence it indicates. Copying a square is one of three drawing tests used in the Binet scale, the others being the diamond (year VII), and the designs to be copied from memory (year X). These tests do not to any great extent test what is usually known as "drawing ability." Only the square and the diamond tests are strictly comparable with one another, the other having a psychologically different purpose. In none of them does success seem to depend very much on the amount of previous instruction in drawing. To copy a figure like a square or a diamond requires first of all an appreciation of spacial relationships. The figure must be perceived as a whole, not simply as a group of meaningless lines. In the second place, success depends upon the ability to use the visual impression in guiding a rather complex set of motor coördinations. The latter is perhaps the main difficulty, and is one which is not fully overcome, at least for complicated movements, until well toward adult life.