ment if the latter is responsible for maladjustment. First, the home situation should be adjusted if possible through education of the parents and child; second, keeping the child at home has the advantage of assisting him to form the habit of facing trying situations; third, removing the child has the disadvantage possibly found in a subsequent return to life on a lower economic scale where a second readjustment is necessary.

R. H. WHEELER (Oregon)

91. STEARNS, A. W., Suicide in Massachusetts. Ment. Hyg., 1921, 5, 752-777

This is a suggestive statistical and qualitative treatment of the problem. One third of the 167 cases studied were definitely insane, while most of the others showed some limitation of responsibility. Since 1841 the suicide rate in Massachusetts has steadily increased. Tables are presented which show that the divorce rate has increased from 26 to 60 per 100,000 since 1870, that suicide is associated only roughly with unemployment, that there are almost four times as many male suicides as female, that for the most part suicides are confined to the relatively poorly educated as far as schooling goes, that the majority of suicides are comfortably situated, economically, that about one half of the suicides are married, that suicides increase with advancing age and that the early spring and summer months show the highest rates. Suicide, he believes, is an expression of a negative self-feeling and so can be caused by any unpleasant experience. The causes of suicide are psychological rather than economic.

R. H. WHEELER (Oregon)

92. CLARK, M. V., Mental Hygiene and the Public Library. Ment. Hyg., 1921, 5, 791-793.

This brief paper shows the need for a more scientific classification of reading matter pertaining to Mental Hygiene and for published statements of authoritative and useful books and articles on specialized topics in this field.

R. H. WHEELER (Oregon)

93. HALE, D. O., Inadequate Social Examinations in Psychopathic Clinics. Ment. Hyg., 1921, 5, 794-806.

As the title suggests, much of the therapeutic work done in clinics underreaches its aim unless social problems are taken into account. Thus the need of a social worker in connection with a clinic is evident.

R. H. WHEELER (Oregon)

94. KLOPP, H. I., How a State Hospital Coöperated with a University to Meet a Community Need. Amer. J. of Psychiat., 1921, 1, 159-166.

A State Hospital has two functions, first as a hospital for mental diseases, second as a part of the general scheme of community service. The Allentown State Hospital's first activity as a community center from the clinical standpoint was its connection with a medical school as a teaching hospital. Second, it held clinics for doctors. Third, it established mental clinics in adjacent cities. Fourth, it became associated as a teaching clinic with Lehigh University. This broadened into a connection with the University Extension Summer School course for teachers, the Biology Department of Muhlenberg College, and the Allentown High School Civics classes. It is in the schools that the state can do the most effective work for the promotion of mental health. In 1919 the Pennsylvania Legislature passed an act pertaining to backward children, which has increased the demand for trained teachers for the special classes. Therefore, in the summer of 1920 Lehigh University offered courses in mental hygiene to teachers. As a part of this course they went to the Hospital where they attended lectures and clinics. In the Department of Psychology the students were given courses in mental diagnosis and the giving of psychometric tests, so the University became a center for the teaching of mental hygiene.

I. M. MACLEISH (Boston Psychopathic Hospital)

95. HOAG, E. B., and WILLIAMS, E. H., The Case of J. P. Watson, the Modern Bluebeard. J. of Crim. Law and Crim., 1921, 12, 348-359.

In this article is described in a somewhat popular fashion the case of a most unusual type of criminal. Watson, a man of considerable refinement, culture and unquestioned intelligence, confessed to having, within a period of three years, contracted twentyone or more illegal marriages and murdered at least nine of these wives. Details of these acts were given absolutely without emotional accompaniment and in an impersonal, coolly-exact manner. Paradoxically, he was quite affected lest his real identity come out, and those, who in his earlier days without parents or support had befriended him, learn thereof. No remorse was felt for the acts themselves, however, which he claimed were committed under the influence of a dominating impulse which allowed him no rest until they were accomplished. In youth and previous to this period, he had always been of a normal, kind and considerate type, experiencing none of the tendencies of the usual moral imbecile towards cruelty to children or animals. He now considers this obsession for killing women as a temporary illness from which he is already recovering fast and of which he will soon be completely rid. Diagnosis, admittedly inadequate and controversial, was: congenital psychic inferiority, with sex perversion, to which is possibly added an epileptic type of personality. "Such cases as these demand the fullest and most complete investigation possible and that the penitentiary authorities should leave no opportunity untouched which might shed light on such tremendously important social, medical and psychological problems."

J. P. CURRIE (Boston Psychopathic Hospital)

II. MENTAL DEVELOPMENT IN MAN

96. PRESSEY, S. L., Empiricism versus Formalism in Work with Mental Tests. J. of Philos., 1921, 18, 393-398.

In reply to a criticism made by Ruml the writer of this article states that the issues may be put briefly, being, (1) He contends for a method more empirical in using tests and a more careful use of hypotheses. Theory with reference to normal distribution curve is too much in evidence in obtaining measures so that the validity of the results seems difficult of determination. A formal statistical method has arisen. Under the circumstances it is safer to return to a thoroughly empirical method. (2) Testing is now a technical science. But the writer would be willing to sacrifice or wave considerations of technique and statistics for practical values as for instance the welfare of some child. But practice of a contrary sort is just now common. (3) All hypotheses not yet verified thoroughly must be eliminated from methods to be used. Empirical methods will bring the experimenter closer to an analysis of practical problems.

T. R. GARTH (Texas)

97. KELLEY, T. L., and TERMAN, L. J., Dr. Ruml's Criticism of Mental Test Methods. J. of Philos., 1921, 18, 459-465. No one has assumed linearity of general intelligence in the sense in which Dr. Ruml appears to use the term. Various sorts of tests should be used in measuring different sorts of mental process, but we may combine those tests if the combination affords more interpretation of the facts than the single test. It is something like appraising a mineral deposit, i.e., securing samples from various parts of the deposit. The Binet scale is for finding out the various levels of "diverse intelligences." Dr. Ruml criticizes the assumption of "rectilinear regressions between intelligence and test performance," but anything else than a fair rectilinear regression in such tests is seldom found. Little error has accrued from assuming rectilinearity, though there are exceptions in trade test findings. The ultimate value of an hypothesis does not depend upon its correctness necessarily, but probably upon its fruitfulness. Dr. Ruml does not believe that the intelligence is static but does not support his statement sufficiently. However, much good will come of such criticisms. The young science is probably girding itself for a new advance.

T. R. GARTH (Texas)

98. DILLINGHAM, A., Superior Children-Their School Progress. J. of Educ. Psychol., 1920, 11, 327-347.

The author makes a critical examination of the course usually prescribed for children of high I.Q. Since getting educated means covering certain ground, the child with the high I.Q. should cover the ground at a more rapid rate. "Probably the most unusual method of attaining this short cut is to test the child 'scientifically' and place him in a class with children whose average mental age corresponds with his. Of course, this arrangement quite ignores character and emotional differences due to physical maturity or the lack of it. Intellect is regarded as the determining factor and no difficulty is anticipated from placing the clever baby of ten with an I.Q. of 150 in a class with average 15-year-old adolescent pupils." The wisdom of this procedure is tested by the statistical study of intelligence and school progress of 174 children. Eighteen case histories are given to illustrate the frequent discrepancies between expected and actual accomplishment and the problems which these children of high I.Q. present to the teacher. The conclusion is that character, emotional and physical traits are to be given equal weight with I.Q. in determining the rate of school progress.

A. T. POFFENBERGER (Columbia)

99. STONE, C. W., and COLVIN, C., How to Study as a Source of Motive in Educational Psychology. J. of Educ. Psychol., 1920, 11, 348-354

This paper affords an interesting illustration of the influence of a motive and of the practice method in increasing efficiency in college classes. Interest in "How to Study" increased rate of reading and degree of comprehension 180 per cent. as shown by comparison with a control group of students. Other factors in the study process showed quite large improvements. It is equally as startling to discover on how low a plane of ability college students work as to discover how much improvement can be made through the devices described in the paper. Any motive, however extraneous, would seem justified by these results.

A. T. POFFENBERGER (Columbia)

100. PRESSEY, S. L., Two Important Points with Regard to AgeGrade Tables. J. of Educ. Psychol., 1920, 11, 355-360.

The two points in regard to age-grade tables discussed at some length are: (1) For many practical situations, the best single statement of the age-grade situation is in terms not of per cent. of retardation, but of median age per grade. This is both more convenient and more sound than the more customary method. (2) Differences in age-grade distribution, from school to school and from school system to school system, must always be taken into account in using tests. This can best be done in terms of median age per grade.

A. T. POFFENBERGER (Columbia)

101. BRIDGES, J. W., The Correlation between College Grades and the Army Alpha Intelligence Tests. J. of Educ. Psychol., 1920, 11, 361-367.

In this report are presented distribution tables for Alpha score and scholarship, Alpha score and academic grade, also for Alpha score and grade in Agriculture, Arts, Engineering and Education separately. The cases were obtained for study by making a random selection from 5,950 test records made at the Ohio State University. One of the most striking things about the data is the large number of students with high Alpha scores and poor academic records. The coefficient of correlation for the largest group studied is plus .35. This is lower than the figures obtained by other investigators. It is, of course, an indication of the importance of the socalled character qualities in academic success. Considerable differences in size of the coefficients for the different schools (Agriculture, etc.) suggest to the author the need for specialized tests for students of the different colleges rather than that Alpha has a selective value in this respect.

A. T. POFFENBERGER (Columbia)