lines of investigation are necessary, and they should not conflict, but should supplement each other. But as to the problem of the appearance of consciousness in the world, we must continue to say: Ignorabimus.

The Measurement of Mental Traits.'

J. CARLETON BELL.

Those interested in the scientific study of education will welcome this book as a contribution to its methods. Professor THORNDike, following the line of GALTON's famous researches, undertakes to bring together in a brief space such of the reliable statistical methods as have already proved fruitful or promise fruitful results in the study of mental traits, especially of school children and college students. Perhaps the greatest value of the book, as the author himself foresees, will be to bring home to educators, more forcibly than heretofore, the untrustworthy character of the current generalizations on education, and to create a demand for inductive statistical study of the facts, such as those being carried on by Professor CATTELL, himself, and others, in our own country.

The implied assumption underlying the whole treatment in this book is that of the possibility of mental measurement This does not mean an attempt to measure that timeless and spaceless, that incommensurable, abstraction that often goes by the name of the "mental" in discussions of the mind-matter problem. It means measurement of the behavior of an organism in terms of those reactions which have come to be called mental because of their relations to the so-called higher values in life-but essentially identical in principle with physical or medical measurements. It is in this sense, apparently, that the author seeks "units of mental measurement" (p. 169), comparable to the inch, the ounce, the ohm, the ampere, the calorie, etc., in physical science. The difficulty is a practical one only. There is no inherent theoretical reason why such a unit may not be found and used. The variability of mental traits renders measurements approximate only. But this is true ultimately of all measurements; they are all anthropic at first. And approximate accuracy is better than the vagaries of current theory, while "the greater the number of measurements, the closer the approximation will be."

If education is to become a science, the physical and mental meas

1 E. L. THORNDIKE, Educational Psychology (Library of Psychology and Scientific Methods, edited by J. MCKEEN CATTELL). New York. Lemcke and Büchner, 1903.

We

urements upon which its conclusions are based must be exact. have a body of general but inexact knowledge about instincts, habits, memory, attention, interest, reasoning. We have descriptions of these in the literature of child-study and methods of teaching. We have a great many general ideas about the influence of inheritance, environment and general mental development. But we have little or no accurate knowledge on these points.

Not much value is attributed to those "Broader Studies of Human Nature" (Chapter XIV), carried on mainly by the questionnaire method, which have emanated from the Clark University school of child-study. The possibility of a scientific study of the loves and hates, fears, interests, ideals, habits, motives and opinions, influence of books, games, toys, etc., is not denied but doubt is cast upon the accuracy of the methods used and the reliability and importance of the generalizations. "Information about 1000 people with respect to one trait is of far less importance than information about 100 traits in each of 10 individuals" (p. 161).

After discussing the possibility of mental measurement (Chapter II), the author takes up the problem of "The Distribution of Mental Traits" (Chapter III). Is there any law of distribution of mental traits in groups of individuals? As between the sexes? As between groups having different inheritance or different training? Can we treat courage, honesty, ambition, eminence, as we can treat color of eyes or hair or weight, statistically? The reply is in the affirmative. But we must beware of imagining "that nature has provided distinct classes corresponding to our distinct words, e. g., normal and abnormal, ordinary and exceptional," genius and idiot, precocious and retarded, bright and dull, etc. (p. 22).

In Chapter IV we have a discussion of "The Relationships Between Mental Traits." Alteration of one function involves others. Educational problems involving this principle are the question of the disciplinary value of studies, arrangement of groups of electives, systems of grading and promotion, tests of mental growth and condition. The relationships are often very different from what the educational literature would have us believe. "The striking thing is the comparative independence of different mental functions even where to the abstract psychological thinker they have seemed nearly identical" (p. 28). The mind is a dynamic, organic, functional whole; not a mechanical whole. It is like the nervous system-a hierarchy of relatively independent activities, "a collection of protoplasmic bands." We have memories not memory, specific habits of attention not a general faculty

of attention; reason is a name for a host of particular capacities. "It follows that an individual's status in any one function need not be symptomatic of his status in others." Hence the fallacy of college entrance examinations as accurate measure of mental traits, and the folly of using any one study, such as arithmetic, as the basis of promotion. In Chapters V and VI Professor THORNDIKE discusses the important questions of "Original and Acquired Traits," and "The Inheritance of Mental Traits." "What ancestry does is to reduce the variability of the offspring and determine the point about which they do vary" (p. 48). There is no theoretical reason why we may not measure the variation and inheritance which determines family resemblance. The author starts, of course, from the work of GALTON, and discusses the small amount of really scientific work which has been done in this field. He does not mention the recent work on MENDEL'S law, which certainly has a bearing. Nor is there any reference to the doctrine of organic selection of OSBORN, MORGAN and BALDWIN as offering a possibility of mediation between the extreme views of the transmissionists and the non-transmissionists.

Chapter VII is on "The Influence of the Environment." Here, again, it is perfectly possible to measure the influence of change in climate, food, school-training, friendship, sermon, occupation, etc. But we must avoid the fallacy here "of attributing to training facts which are really due to original nature or selection." The author would substitute for such vague and indefinite terms as culture, discipline, training, practice, imitation, the conceptions of "(1) Furnishing or withholding conditions for the brain's growth and actions; (2) Furnishing or withholding adequate stimuli to arouse the action of which the brain is by original nature or previous action capable; (3) Reinforcing some and eliminating others of those activities in consequence of the general law of selection in mental life" (p. 77).

One of the most valuable Chapters in the book is Chapter VIII on "The Influence of Special Forms of Training Upon General Abilities." "Does the study of Latin or of mathematics improve one's general reasoning powers? Does laboratory work in science train the power of observation for all sorts of facts?" In other words, "How far does the training of any mental function improve other mental functions ?" (p. 80). There is no doubt that there is some influence. The question is, "To what extent and how" does this take place? "Learning to do one thing well has much less influence upon one's other abilities' than educational theorists would have us think. The general conclusion from his own experiments is "that a change in one function alters any other

only in so far as the two functions have as factors identical elements" (p. 80). "Improvement in any single mental function need not improve the ability in functions commonly called by the same name. It may injure it" (p. 91). There is no "general ability." Upon this the author repeatedly insists. The present reviewer thinks that Professor THORNDIKE carries his idea of the independence of the mental functions to a point which threatens the unity of the mental life. One wonders how a mind such as the author describes ever could perform such a synthesis as that involved, for example, in writing book on Educational Psychology. He says that "the mind must be regarded not as a functional unit nor even as a collection of a few general faculties which work irrespective of particular material, but rather as a multitude of functions, each of which is related closely to only a few of its fellows" (p. 29). "The mind is really but the sum total of an individual's feelings and acts" (p. 30). "This view is in harmony with what we know about the structure and mode of action of the nervous system. The nervous system is a multitude of connections between particular happenings in the sense organs and other particular events in the muscles' (p. 30).

These unguarded statements surely must be accounted for as the result of a violent recoil from the extremes to which the "abstract psychological thinker" has carried the faculty psychology. It cannot be that Professor THORNDIKE means to deny the important structural and functional unities found in the nervous system and in conscious pro

cess.

Chapter X treats of "Changes in Mental Traits with Age,' and Chapter XI of "Sex Differences in Mental Traits." No mention is made of Professor HELEN BRADFORD THOMPSON'S recent work on "Psychological Norms in Men and Women." Chapter XII is on "Exceptional Children," especially defective children. A brief concluding chapter puts the "Problem of Education as a Science." An Appendix contains an "Index of Tests," of "Common Measures," and "Suggestions for Investigations in Educational Science."

The author is rather cavalier in his treatment of educational theory. But most of his readers will probably forgive him for that. As before remarked, the book is chiefly valuable as setting the task and suggesting the methods of a scientific study of education. It can scarcely be said that it adds much of positive value in the way of conclusions from data already studied. There are very few of the generalizations contained in this book which it would be safe to adopt without further vindication of their truth. But it certainly will stimulate more

exact methods of study of mental traits in relation to education, and this is more than justification for its appearance at this time.

H. HEATH BAWDEN.

Are Sounds, Made in the Air, Audible in the Water?

In their works on the sense of hearing in fishes and crustacea, KREIDL, BEER and PARKER have laid more or less emphasis on the reflection of sound waves in the air from the surface of the water. PARKER says (basing his statement to some extent upon experiment), "the plane separating air and water is, under ordinary circumstances, an almost impenetrable one for most sounds, whether they are generated on one side or the other of it, and many of the negative results obtained by previous investigators on the sense of hearing in fishes may have been due not so much to the absence of hearing in the animals experimented upon as to their inaccessibility to the sound, or at least to sound of an intensity sufficient to stimulate."

Interesting experimental evidence on this question is furnished by Dr. V. DUCCESCHI,' of Naples, in a recent number of the Rivista d'Italia. Struck by the fact that some boys, diving along the shore, were able to repeat, on emerging from the water, the words called to them by their comrades while they were still beneath the surface, he secured the services of an expert diver, provided himself with a boat and some simple apparatus, and set out to test the matter experimentally. Trials were made at various depths up to seven meters. The length of time that the diver remained under water was about 10 seconds. At 5 meters the diver could hear distinctly, and repeat on coming to the surface, every word called to him from the boat. At 6 meters he could distinguish between the sounds of two glass bells of different sizes, a whistle and a small trumpet, all sounded in the air, could tell how many times each one was sounded, and in what order. At 7 meters the diver was able to distinguish the sounds with much less certainty, and sometimes not at all. The high tones were found to be much more difficult to distinguish than the low.

The same set of experiments was tried when the water was somewhat rough, with the result that the sounds were perceived with slightly less accuracy. It is true that in all these experiments the possibility of the sound being communicated through the boat to the water is not excluded. Moreover DUCCESCHI thinks it may be a question whether

IV. DUCCESCHI. Gli animali aquatici possiedono il senso dell 'udito? Rivista d'Italia, Anno VI, pp. 958-966, Dec., 1903.