This little experiment, simple as it is, seems to fulfill reasonably well the requirements of a good test. The main objection which might be brought against it is that it is much subject to the influence of training. If this were true in any marked degree, the mentally retarded children of 7-year intelligence should be expected to succeed better with it than mentally advanced children of the same mental level, since the former would have had at least two or three years more in which to learn the task. A comparison of the two groups, however, shows no great difference. The factor of age, apart from mental age, affects the results so little that it is evident we have here a real test of intelligence. It would, of course, be easy to imagine a child of 7 years who had not had reasonable opportunity to make the acquaintance of bow-knots or to learn to tie them. But such children are seldom encountered in the ages above 6 or 7. Of 68 7-year-olds who were asked whether they had ever seen a bow-knot ("a knot like that") only two replied in the negative. It cannot be denied, however, that specific instruction and special stimulus to practice do play a certain part. This is suggested by the fact that girls excel the boys somewhat at each age, doubtless because bow-knots play a larger rôle in feminine apparel. Social status affects the results in only a moderate degree, though it might be supposed that poor ragamuffins, on the one hand, and children of the very rich, on the other, would both make a poor showing in this test; the former because of their scanty apparel, the latter because they sometimes have servants to dress them. The following are probably the chief factors determining success with this test: (1) Interest in common objective things; (2) ability to form permanent associative connections between successive motor coördinations (memory for a series of acts); and (3) skill in the acquisition of voluntary motor control. The last factor is probably much less important than the other two. Motor awkwardness often prolongs the time from the usual ten or fifteen seconds to thirty or forty seconds, but it is rarely a cause of a failure. The important thing is to be able to reproduce the appropriate succession of acts, acts which nearly all children of 7 years, under the joint stimulus of example and spontaneous interest, have before performed or tried to perform. VII, 5. GIVING DIFFERENCES FROM MEMORY