A few critics have objected to the use of tests demanding abstract thinking, on the ground that abstract thought is a very special aspect of intelligence and that facility in it depends almost entirely on occupational habits and the accidents of education. Some have even gone so far as to say that we are not justified, on the basis of any number of such tests, in pronouncing a subject backward or defective. It is supposed that a subject who has no capacity in the use of abstract ideas may nevertheless have excellent intelligence "along other lines." In such cases, it is said, we should not penalize the subject for his failures in handling abstractions, but substitute, instead, tests requiring motor coördination and the manipulation of things, tests in which the supposedly dull child often succeeds fairly well. From the psychological point of view, such a proposal is naïvely unpsychological. It is in the very essence of the higher thought processes to be conceptual and abstract. What the above proposal amounts to is, that if the subject is not capable of the more complex and strictly human type of thinking, we should ignore this fact and estimate his intelligence entirely on the ability he displays to carry on mental operations of a more simple and primitive kind. This would be like asking the physician to ignore the diseased parts of his patient's body and to base his diagnosis on an examination of the organs which are sound! The present test throws light in an interesting way on the integrity of the critical faculty. Some subjects are unwilling to extend the report in the least beyond what they know to be approximately correct, while others with defective powers of auto-criticism manufacture a report which draws heavily on the imagination, perhaps continuing in garrulous fashion as long as they can think of anything having the remotest connection with any thought in the selection. We have included, for each selection, one illustration of this type in the sample failures given above. The worst fault of the test is its susceptibility to the influence of schooling. Our uneducated adults of even "superior adult" intelligence often fail, while about two thirds of high-school pupils succeed. The unschooled adults have a marked tendency either to give a summary which is inadequate because of its extreme brevity, or else to give a criticism of the thought which the passage contains. This test first appeared in Binet's 1911 revision, in the adult group. Binet used only selection (b), and in a slightly more difficult form than we have given above. Goddard gives the test like Binet and retains it in the adult group. Kuhlmann locates it in year XV, using only selection (a). On the basis of over 300 tests of adults we find the test too difficult for the "average adult" level, even on the basis of only one success in two trials and when scored on the rather liberal standard above set forth. SUPERIOR ADULT, 5: REPEATING SEVEN DIGITS REVERSED