used in elaborating the dream is mental: the nature of the memories revived being wholly controlled by the nature of the mental stimulus. Possibly the true wish' dream and the true 'physical stimulus' dream are not irreconcilable alternatives but extreme cases of the dream in which both internal mental stimuli and external physiological and physical stimuli act upon the complex retent' in mind and conjure up a jumbled memory having hallucinatory vividness.

In Chapter vi we have the familiar summary of Freud's theory of 'dream work' and in Chapter VII some account of "Rivers's view of the dream." "The undisguised and terrifying dream of battle...the nightmare and the undisguised sexual dream...led Dr Rivers to contest at several significant points Freud's explanation of dreams, more particularly at those concerning the censorship, the alleged sleep preserving function of the dream and the wish fulfilment theory.... As Rivers views it the dream arises out of mental conflict and is an attempt...to solve a problem" (p. 102). With Rivers's criticisms and his doctrine of levels of experience the author appears to be in agreement. He also throws in his lot with those psychologists who deny that "the dream work" is, in any essential, different from activities of the ordinary everyday waking consciousness. He writes: "It is sometimes claimed that the processes of the dream work are unique, that they have no parallels in waking life. This is inadmissible. Freud's view: 'It is condensation that is mainly responsible for the strange impression of the dream, for we know nothing analogous to it in the normal psychic life accessible to consciousness,' is expressed too absolutely" (p. 109). He then argues that in the 'generic' image, portmanteau-words' and the cartoon we have waking examples of condensation; that dramatization is common in waking life: "many visualizers attempt the solution of almost all difficulties which require forethought by picturing different actions and their results" (p. 114); that "secondary elaboration is only the customary manner of interpreting any object or thought which has been imperfectly apprehended" (p. 114), a fact that has been demonstrated over and over again in experimental work on perception and on memory, and concludes "the dream, therefore appears to be a mental structure the constituent mechanisms of which are not different in kind from those which characterize the mental events of waking life. It is the reciprocal interplay of these mechanisms, the altered emphasis which each of them receives, and their comparative freedom from the dominating directive tendencies of the day time which combine to make the dream the enfant terrible of the well ordered personality, and the delight of the modern psychologist."

Coming closer to the problem of remembering and forgetting, in Chapter IX, "How we forget," we find the "older psychology," including the work of Ebbinghaus, treated very briefly and somewhat cavalierly and then the part played by the affect in forgetting is considered at length. Freud's doctrine of repression and Rivers's discrimination between 'witting repression' and 'unwitting suppression' are dealt with, while Rivers's attempt to indicate a physiological explanation of forgetting by analogical comparison with facts and theories associated with epicritic and protopathic sensation, the cerebral control of the optic thalamus and the 'mass-reflex' of the divided spinal cord is favourably reviewed, subject to a note that "the validity of [Rivers's interpretation of the facts] is not granted by all physiologists" and a reference to Metcalf's paper1. Finally, replacing Rivers's idea of the fusion of memories by 1 Psychological Bulletin, 1921, xvI. 4, 181–202

the idea of 'embodiment,' the author puts forward the following as a provisional classification of forgotten experiences:

1. Embodied

2. Exiled

3. Superseded

((a) Apparently insignificant.

(b) Significant but completely congruous with
the personality.

Forgetting of Class 1 (a) "may conceivably be due to physiological decay" (p. 166), but if all experience is indefinitely retained this explanation must be rejected. The author however is a little sceptical of the doctrine of complete retention and writes "in a collection of my own dreams I have found memories of very early experiences of childhood...but I have not discovered any whichdid not prove to be part of the associative fringe of some very significant incidents. And it may conceivably prove to be true that only those incidents which for some reason are disembodied... form the material for the hypnotist's striking performances" (p. 167). In Class 1 (b) we have the "obliterating effect of congruity upon past experience." Class 2 comprises the 'repressed' experiences of psychopathology and Class 3 obsolete memories that "do not appear to be held out of consciousness by an ever present resistance...[but] are seldom invited to enter" (p. 174).

Except in Chapter 1 and portions of Chapter IX the author has written of imagery and the dream rather than of memory. The imagery used in recall is of far less importance in the study of memory than the reliability and amount of the recall, while to classify the dream as memory would seem an error unless 'memory' is to be made co-terminous with thought. There is a sense in which the dream is memory, but in exactly the same sense the falsehood "I visited Sirius last week and discussed the matter with Julius Caesar" is memory. The author has expressed it thus: "The statement [that it is memory] is true rather of the material and the constituent patterns than of the main design of the dream" (p. 69). We would suggest that it is essential that classification of the dream should not exclude this main design. If Freud is right and dreams are 'wish' fulfilment they would seem to rank with the free associative, imaginative thought of 'day-dreams' and 'castles in the air.' If Rivers is right and they are attempts to solve a problem they would seem to rank with the directed, selective thought of constructive imagination. If, further, there are dreams that fit neither theory they would seem far more closely allied to hallucination than to either the free associative thought of reminiscence or the controlled, critical thought of recollection. Memory is tinged with a temporal signature that places the actuality in the past: the temporal signature of the dream is now. As an experience the dream has more in common with the percept than with memory, but its relation to reality is not that of the percept. As its relation to time is not that of memory and its relation to reality is not that of the percept, the dream can scarcely be classified as either the one or the other, and would seem to be much better placed either as ballucination or as imagination. The author was very close to this view when, in discussing the processes of dream-work, he wrote: "The evidence, indeed, makes it extremely probable that they are operative not only in fashioning the dream but in the work of the waking activity of creative imagination" (p. 87).

R. J. BARTLETT.

Group Tests of Intelligence. By PHILIP BOSWOOD BALLARD, M.A., D.Lit., London. Hodder and Stoughton, 1922. pp. x + 252.

Dr Ballard is one of those who have firmly grasped the indubitable fact that "intelligence tests" have come to stay. Once or twice in the course of a generation there emerges in the conduct of human affairs a factor which, in spite of all hostility and suspicion, calmly and inevitably developes, leading eventually to a complete overhauling of the principles and methods applicable in the field to which it belongs. The procedure of testing intelligence on the basis of age-performance by methods such as those described by Dr Ballard in this book (and in his previous book, Mental Tests) is such a factor. There is little doubt that it is destined ultimately to revolutionize our ideas on education and on the fitting of the child to bear his share in the national life in the position most appropriate to him by reason of the type and grade of his native mental ability.

Criticisms of intelligence tests are mainly academic and a priori in kind. They are usually based on the appearance of the tests, and not on their results. Yet it is their results as empirically observed, which determine for the psychologist the applicability and suitability of his tests, and by their results the tests must be judged. The only thing which really matters is whether the tests do in fact perform successfully their task of grading children (and others) accurately according to their capacity for being educated and for applying what they have gained by education. Though by no means perfect, the tests are already performing this task well (far better, indeed, than any other method of the past) and, as Dr Ballard points out, they are performing it better every dav.

The book under consideration falls into four well-defined sections. Dr Ballard first gives an account of the standard types of group tests derived from the American experiments (notably the wholesale testing of the American Army). He then passes on to a general survey of group testing in England, and in particular, of the types of tests which he himself has used. There follows a discussion of the nature and limits of intelligence. Finally, there is a valuable section dealing with the elements of the statistical technique necessary to a proper collation and interpretation of the results of the tests.

Dr Ballard's account of the development in England and America of methods of measuring intelligence is clear and interesting, and forms what is perhaps the most useful summary of these methods at present obtainable. The chapter on Dr Godfrey Thomson's well-known "Northumberland Tests" will be of particular interest to English readers; while the account of the author's own tests provides valuable and suggestive additions to the armoury of diagnostic weapons now available.

The chapter on the nature of intelligence affords material for what will always be a source of keen discussion. As Dr Ballard makes clear, the question is, of course, mainly of academic rather than practical interest so far as "intelligence tests are concerned. All who are familiar with the practice of them will agree that the tests measure something, but the ability to frame a precise definition of this "something" is irrelevant to the question of the practical value of the tests, which can be decided only by their capacity to perform the task to which attention has been previously drawn, namely the selection of children according to their educability.

With Dr Ballard's remarks on the limits of the growth of intelligence the

present writer is in full agreement. It has always been a source of much surprise to him that so many people find great difficulty in accepting the now demonstrable conclusion that the growth of "intelligence" (in the significant, if not precise, commonsense meaning of that term) ceases at a comparatively early age. He suspects the existence of a "complex" lurking at the root of this difficulty. But, in any case, he can see no reason, quite apart from the experimental evidence, for the alleged a priori difficulty in supposing that we do not get any better at reasoning as we grow older. The simple fact seems to be that, through acquired experience, we get more data, as we grow older, on which to base our reasoning, and are therefore able to meet more successfully the various situations with which we are faced.

The closing chapters of the book on the statistics of correlation cannot fail to be useful. For there are many hidden perils awaiting the mental tester when he comes to interpret and apply his results; and if the latter are to lead to a successful re-organization of educational methods on lines similar to those briefly considered by Dr Ballard in his final chapter, it is of the first importance that experimenters should be warned of these perils in advance, in order that they may exercise the vigilant and critical scrutiny necessary to avoid them. C. A. RICHARDSON.

Methods and Experiments in Mental Tests. By C. A. RICHARDSON, M.A. London: George Harrap & Co., 1922. pp. 94. Price 3s. 6d. net.

This small but useful book does not purport to give a general account of mental tests. There is a brief introductory reply to some of the common but ill-informed criticisms of tests, but the book is mainly concerned with a discussion, on the basis of the author's own experiments, of the reliability of the Stanford-Binet scale as an index of educable capacity, the derivation of mental age from scores in a group test, methods of estimating the 'true' intelligence quotient of adults and adolescents, and the reliability of the group intelligence test as an index of educability. These problems, and the research material which the author has to contribute, are clearly and directly set out, without any waste of words, and in such a way as to be readily understood and appreciated by the interested but non-technical reader for whom the book is intended.

As evidence of the value of the Stanford-Binet scale as an index of educable capacity, the results of an investigation into the correlations between the intelligence quotient and attainment in arithmetic and composition with five groups of twenty children are given. The correlations were high, and an analysis of the deviations showed them to be due to not more than about 15 per cent. of the children, with assignable causes. For the whole group of 100 children the degree of scatter of the E.Q. (educability quotient) was markedly less than that of the I.Q., probably indicating insufficient elasticity of promotion for the brighter children. The suggestion is made that the S.B. scale might well be applied to children twice during their school life, namely at seven and eleven years of age.

For the derivation of the mental age of individual children from a group test score (which in itself yields nothing more than the relative intelligence of the members of the group), the formula y = x + 110 is offered, where y is mental age in months, as estimated by the S.B. scale, and x the score

in the group test. It is considered, the author's reasons being clearly shown, that this formula gives with some accuracy the correspondence in general between true mental age and score in (Terman) group test; but it is pointed out that errors may occur in individual cases mainly because the two tests do not cover exactly the same ground, the group tests so far standardized not being yet sufficiently comprehensive.

The factors in the problem of estimating the true intelligence quotients of adults and adolescents are well shown. Two methods of estimating the I.Q. in these cases are suggested, one of which, a percentile rank method, was set out in a note in the British Journal of Psychology, April 1922. The other method works with the formula

where the effective mental age is the age at which the score made by a subject would be reached by him if his intelligence continued to grow indefinitely at the same rate as in childhood, and the effective age of a subject is the age at which the average score corresponding to his actual age would be reached if intelligence continued to grow at the same rate as in childhood.

The reliability of the group test as an index of the quotient of educability is approached by the indirect method of comparing the results of group and individual intelligence tests, with conclusions already noted; and by the direct method of comparing the results of a group test with those of an ordinary written examination in arithmetic and English. This comparison was made with some 500 children, and, allowing for certain defects in technique, the correspondence was very close, strikingly so in a considerable proportion of cases, confirming the general conclusion that tests of the group scale type are useful and sufficiently accurate means of estimating educability. But "this reliability is likely to be increased when our tests are so devised as to probe the child's intelligence from as many directions as possible."

Whether the detailed methods and results of the author stand or fall, there can be no doubt that the book is of much interest and importance to those who are becoming aware of the educational significance of mental tests. Perhaps there is too little reference to the work of other investigators; the chief value of the book undoubtedly lies in its admirable exposition of the character of the problems arising out of the practical application of mental tests, and of the methods by which these are to be approached.

S. S. BRIERLEY.

The Psychology of Self-Consciousness. By JULIA TURNER, B.A. (Lond.). Kegan Paul, Trench, Trubner & Co., Ltd. pp. xii + 243. Price 6s. 6d. net. The author has given us a very interesting hypothesis of the development of self-consciousness as a resultant of life-hunger and fear when confronted by an awesome 'Not-I.' These blend to give anxiety, which not only creates, but dominates human life. For sanity and health the 'power sense,' or 'will to live' must balance the 'expiation tendency,' or conscious fear of and desire to propitiate superior agencies. To the reviewer these seem to correspond closely to McDougall's instincts of self-assertion and self-abasement, which, under analysis, are found to arise at the level of self-consciousness in much the same way that the author has suggested as probable. They do not, however,