(•:' ■( * ^ "O & c^ fV Y ft V^^ ' * ft ^ _-?^ '=%,«^ - .^^ .^^ °^- "■--%.^*- cP^-i'/''„% cp^^.i-./^ cr.-;v',% cT-'-^"'' ' A*^ ^^ V CP -^' *"' V^ ^^ A*^ <* ''/?ls^ A*^ <- "/7. aO^ ^ "/..^^ ^A^ <* A*-"" <^ •/,-,. s ~ A^ <*■'<• ^ ^ X' "' A^ ^ •.r\^ ;^.^^ •^ >^'' '>' % ^ "'',., s^ A^ <'''*'«'' ^ A*^ ^ ^ H A" 1' ,& rb The Psychology of the Negro AN EXPERIMENTAL STUDY GEORGE OSCAR FERGUSON. A. M. RKPRINT OF ARCHIVES OF PSYCHOLOGY No. 36. Api»IL. X016 PSVCHOLOOT. VOLUMK X'XV, ISTO. 1 SUBMITTISD IN PARTIAL FUXJEnLMENT OF" THK WRQaiKKArENTS FOR mXR DEGREE OF DOCTOR OF PHTIiOSOPHY IN THE FAOUI/TY OF PHILOSOPHY CX)IitJMBIA UNIVERSITY NBJW YORK 1016 The Psychology of the Negro AN EXPERIMENTAL STUDY BY GEORGE OSCAR FERGUSON. Jr.. A. M. REPRINT OF ARCHIVES OF PSYCHOLOGY No. 36, April, 1916 Columbia University Contributions to Philosophy ano PeYCHOi.OQY, Volume XXV, No. 1 SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHIIiOSOPHY IN THE FACULTY OF PHILOSOPHY COLUMBIA UNIVERSITY NEW YORK 1916 Ifl? CONTENTS CHAPTER PAGE I. Review of Work Previously Done 1 Non-Experimental Studies, 2. Boas, 2; Le Bon, 4; Tylor, 4; Hall, 5; Odum, 7; Bardin, 8; Galton, 8; Thorndike, 8, Experimental Studies, 9. Bache, 9; Smith, 10; Stetson, 10; McDonald, 10; Wood- worth, 11; Mayo, 13; Phillips, 14; Strong, 17; Baldwin, 19; Pyle, 20. Neurological Studies, 22. Bean, 22; Mall, 23; Hrdlicka, 24; Summary of Chapter 25. II. The Subjects and the Tests 27 The Subjects, 27. White and Negro Population and Education in the Localities Tested, 27; Relative Action of Selective Factors upon Whites and Negroes in Schools, 28; Number and Distribution of Subjects Tested, 32; Com- parative Ages of Whites and Negroes Tested, 35. The Tests, 37. Mixed Relations Test, 37; Completion Test, 39; Maze Test, 41; Cancellation Test, 41; Method of Giving the Tests, 41; Method of Scoring, 42; Reliability of Short Tests, 45. III. General Comparison of Whites and Negroes. . . 46 Treatment of Data, 46. Mixed Relations Test, 50. Relative Ability of Pupils of the Same Age but Dif- ferent Grades, 55. Completion Test, 61. Maze Test, 66. Cancellation Test, 75. Relative Variability of Speed and Accuracy in Tests, 79; Racial Sex Differences, 83. IV. Comparison of Sub-Classes of Negroes 84 Number and Distribution of Pure Negroes and Mulat- toes in the United States, 84. Views of other Writers, 87. Boas, 87; Hall, 88; Jordan, 88; Le Bon, 89; Baker, 89; Stone, 90. The Classification, 90. Numbers and Ages, 92. Standing in the Tests, 95. Treatment of Data, 95; Mixed Relations Test, 95; Completion Test, 105; Relation Between Skin Color and Racial Purity, 110. Comparative Variabilities, 111. Significance of Variability, 113; Methods of Measur- ing Variability, 113; Results, 114. V. Conclusion 123 General Summary of Conclusions, 123; Ability of Mulattoes, 125; Educational Considerations, 125; Com- parison of Whites and Negroes in Terms of the Num- ber of Eminent Men to be Expected from Each Race, 127; Considerations Bearing upon the Future of the Negro, 128; Death Rate of Whites and Negroes, 131. References 133 Appendix 136 AUTHOR'S NOTE The author wishes to express his thanks To Superintendents J. A. C. Chandler, of Richmond, and E. F. Birckhead, of Fredericksburg; to Acting-Superintendent E. W. Huffman, of Newport News; to Principals J. H. Brent, S. D. Turner, W. G. Jones and J. C. Harwood, of Richmond, W. L. Ransone, of Fredericksburg, and D. A. Dutrow, of Newport News ; and to the teachers whose grades were tested, for their cooperation in this study; To Professors J. McK. Cattell and E. L. Thorndike, and Dr. A. T. Poffenberger, Jr., of Columbia University, for their helpful suggestions; To Colonel Clarence Hodson, of Newark, for his thoughtful interest and kind assistance in the preparation of the work; To Professor R. S. Woodworth, of Columbia University, for his constant readiness to aid, his helpfulness in many ways during the course of the research, and his many kind- nesses. THE PSYCHOLOGY OF THE NEGRO- AN EXPERIMENTAL STUDY CHAPTER I REVIEW OF WORK PREVIOUSLY DONE Interest in the psychology of the negro has produced a voluminous Hterature, but the knowledge to be obtained from a reading of it is not commensurately extensive. It may be not unjustly said that until what is practically the present time our information as to the negro's intellectual character- istics has been almost wholly a product of varying individual opinion and speculation. Here and there have appeared works of value, based upon study and experience and presenting carefully drawn conclusions. But for the most part the litera- ture consists of articles which have grown out of limited and untrustworthy observation, and of articles which have attacked the problem from the standpoint of preconceived theories and have reached conclusions a priori from the premises thus held. There has been no settled body of doctrine concerning the vastly important matter of the mental capacity of the negro. One man has held that the negro is the equal of the white in intellect; another has held that a great intellectual gulf sep- arates the two races. And there have been many varieties of views between these two extremes. There have been no facts agreed upon and consequently no reliable generalizations. Yet social practices of far-reaching importance have been based upon these varying views. Some school systems have advo- cated giving precisely the same training to precisely the same racial minds; other systems have advocated a differentiation of school work to meet the needs of two mentally different races ; the advocates of both views have put their beliefs into 2 THE PSYCHOLOGY OF THE NEGRO. practice. Many social and political considerations have of course had their bearing upon these educational matters, but certainly ideas as to the nature of the mind of the negro have not been without influence. And the social and political con- siderations themselves have had a psychological background. In the last few years a number of objective studies of the intellect of the negro have been made, and they constitute a definite step toward a scientific answer to the vexed question upon which they bear. It is the purpose of the present chapter to review the experimental work which has been done in this limited field of race psychology and also some of that which has not been experimental. In the following chapters will be set forth the results of an objective study which it is hoped will contribute in some measure toward an answer to the problem. Non-Experimental Studies In discussing "human faculty as determined by race," Boas, in an early article ('94), the substance of which is in- corporated in a later work ('11), pointed out that while the skull capacity of modern European whites is 1560 cc. and that of European whites of the neolithic period is the same, the skull capacity of the mongaloid race is 1510 cc, that of the negroes of the Pacific ocean is 1460 cc. and that of African negroes is 1405 cc. The negroes, too, were at least as tall as Europeans. Another way of putting it is to say that while 50 per cent, of whites have skull capacities of 1560 cc, only 27 per cent, of negroes equal or exceed this skull capacity. Further, "We find that the face of the negro as compared to the skull is larger than that of the American, whose face is in turn larger than that of the white. The lower portion of the face assumes larger dimensions. The aveolar arch is pushed forward and thus gains an appearance which reminds us of the higher apes. There is no denying that this feature is a most constant character of the black races and that it rep- resents a type slightly nearer the animal than the European type." ('94,p. 311). "Our conclusion is, that there are differ- ences between the physical characters of races which make it probable that there may be differences in faculty. No un- questionable fact, however, has been found yet which would prove beyond a doubt that it will be impossible for certain races to attain a higher civilization." ('94, p. 317). In a later article ('01, p. 3), Boas argues as follows: "A REVIEW OF WORK PREVIOUSLY DONE. 3 number of anatomical facts point to the conclusion that the races of Africa, Australia and Melanasia are to a certain extent inferior to the races of Asia, America and Europe. We find that on the average the size of the brain of the negroid races is less than the size of the brain of other races, and the differ- ence in favor of the Mongaloid and white races is so great that we are justified in assuming a certain correlation be- tween their mental ability and the increased size of their brains. At the same time it must be borne in mind that the variability of the mongaloid and white races on the one hand, and of the negroid races on the other, is so great that only a small number, comparatively speaking, of the individuals belonging to the latter have brains smaller than any brains found among the former; and that, on the other hand, only a few individuals of the mongaloid races have brains so large that they would not occur at all among the black races. That is to say, the bulk of the two groups of races have brains of the same capacities, but individuals with heavy brains are proportionately more frequent among the mongaloid and white races than among the negroid races. Probably this differ- ence in size of the brain is accompanied by differences in structure, although no satisfactory information on this point is available." Boas then takes up the argumpnts that primitive races cannot abstract, inhibit impulses or choose according to standards of value. His contention is that primitive man does do tnese things, but that he does them from his own pomt of view and to meet his own needs, and not in the same way that civilized man does them. Similarly with the argument that while a savage can perceive well in a sensory way, he cannot interpret phenomena: he does interpret phenomena, but from his own point of view, as is the case with all men. "Our considerations make it probable that the wide differences between the manifestations of the human mind in various stages of culture may be due almost entirely to the form of individual experience, which is determined by the geographical and social environment of the individual. It would seem that, in the different races, the organization of the mind is on the whole alike, and that the varieties of mind found in different races do not exceed, perhaps not even reach, the amount of normal individual variation in each race." ('01, p. 11). A radically different opinion from that of Boas is held by 4 ', THE PSYCHOLOGY OF THE NEGRO. Le Bon. ('98). This author holds that the human race may be divided into four groups on the basis of psychological char- acteristics: (1) Primitive races, such as the Fuegians and the aboriginal Australians, (2) Inferior races, such as the negroes, (3) Average races, such as the Chinese, Japanese, Mongolians and Semitic peoples, (4) Superior races, which are the Indo-Europeans. "No confusion is possible between the four great divisions we have just enumerated. The men- tal abyss that separates them is evident." ('98, p. 28) . The specific differences which separate the primitive and inferior peoples from those which are higher are that the former races have a relative incapacity to reason or associate, to com- pare and draw conclusions, to attend, observe and reflect, to exercise foresight, to persist in a given line of activity, to hold to a distant rather than a present end. These differences are practically ineradicable, and they determine the achievement of the races. "The various elements of the civilization of a people being only the outward signs of its mental constitution, the expression of certain modes of feeling and thinking peculiar to a people, these elements cannot be transmitted unchanged to peoples of a different mental constitution: all that can be transmitted is the exterior, superficial, and unimportant forms." ('98, p. 233). Races of men also differ in the relative size of their brains, according to Le Bon. From measurements of the volume of several thousand skulls, he concludes that the differences are real, though not very considerable. And he further draws the interesting conclusion that the more civilized races display a much greater divergence from their average brain size than do the races which are backward. Thus the higher races have more very large and very small brains than primitive people, relatively to their average; the brains of the inferior races conform more nearly to their average type. Tylor writes ('04, pp. 74-75) : "There seems to be in man- kind inbred temperament and inbred capacity of mind In measuring the minds of the lower races, a good test is how far their children are able to take a civilized education. The account generally given by European teachers who have had the children of lower races in their schools is that, though they often learn as well as the white children up to about twelve years old, they then fall off, and are left behind by the children of the ruhng race. This fits with what anatomy REVIEW OF WORK PREVIOUSLY DONE. 5 teaches of the less developed brain in the Australian and Afri- can than in the European. It agrees also with what the his- tory of civilization teaches, that up to a certain point savages and barbarians are like what our ancestors were and our peasants still are, but from this common level the superior intellect of the progressive races has raised their nations to heights of culture." "Moreover, there is this plain difference between low and high races of men, that the dull-minded barbarian has not power of thought enough to come up to the civilized man's best moral standard Much of the wrong-doing of the world comes from want of imagination The lower races of men are so wanting in foresight to resist passion and temptation, that the moral balance of a tribe easily goes wrong, while they are rough and wantonly cruel through want of intelligent sympathy with the sufferings of others, much as children are cruel to animals through not being able to imagine what the creatures feel." ('04, pp. 407-408). In discussing racial brain differences Tylor says : "Professor Flower gives as a mean estimate of the contents of skulls, Australian, seventy-nine; African, eighty-five; European, ninety-one. Eminent anatomists also think that the brain of the European is somewhat more complex in its convolutions than the brain of a Negro or Hottentot. Thus, though these observations are far from perfect, they show a connection between a more full and intricate system of brain-cells and fibres, and a higher intellectual power, in the races which have risen in the scale of civilization. "It is often possible to tell by inspection of a skull what race it belongs to. The narrow cranium of the negro would not be mistaken for the broad cranium of the Samoyed Taking the diameter from back to front as 100, the cross- diameter gives the so-called index of breadth, which is about 70 in the Negro, 80 in the European, and 85 in the Samoyed. The Austrahan and African are prognathous or "forward jawed,' while the European is orthognathous, or 'upright- jawed.' At the same time the Australian and African have more retreating foreheads than the European, to the disadvan- tage of the frontal lobes of their brain as compared with ours." ('04, pp. 60-62). G. Stanley Hall's view of the relative mental make-up of the negro and the white may be set forth in the following 6 THE PSYCHOLOGY OF THE NEGRO. quotations : "No two races in history, taken as a whole, differ so much in their traits, both physical and psychic, as the Caucasian and the African. The color of the skin and the crookedness of the hair are only the outward signs of many far deeper differences, including cranial and thoracic capacity, proportions of body, nervous system, glands and secretions, vita sexualis, food, temperament, disposition, character, longevity, instincts, customs, emotional traits and diseases. All these differences, as they are coming to be better under- stood, are seen to be so great as to qualify if not imperil every inference from one race to another, whether theoretical or practical, so that what is true and good for one is often false and bad for the other." ('05, p. 358). "Another racial trait of the negro is found in the sphere of sexual development. Special studies show that the negro child up to about twelve is quite as bright as the white child ; but when this instinct develops it is earlier, more sudden, and far more likely permanently to retard mental and moral growth than in the white who shoots ahead. Thus the virtues and defects of the negro through life remain largely those of puberty." ('05, p. 362). This last contention of Hall's, that the negro's develop- ment comes to what is at least a partial stand-still at puberty, occurs in the writings of others who have dealt with the sub- ject. Tylor has already been quoted on this point. The idea is that after puberty the individual's mental life broadens and takes on new aspects: abstraction, a tendency to penetrate into the meanings of things, the power to perceive relations, and the ability to appreciate logical, aesthetic and moral situa- tions. Before adolescence a child's activities are mainly the so-called lower mental processes, such as perception, memory and the motor responses. The negro, being the lower type, fails to attain the post-pubertal traits to the degree that the white child attains them, and therefore remains permanently on a lower level. But on this level, and in the traits which constitute it, he is fully the equal of the white child. Thus, while he cannot reach the finer elements of mental attain- ment, the negro is yet the equal or the superior of the white in sense capacity, rote memory, objective attentiveness, motor control, and qualities of a similar nature. Another statement from Hall bears upon this : "Mental development after puberty is much more uncertain than before. The first twelve years REVIEW OF WORK PREVIOUSLY DONE. 7 of life represent larger and more fundamental qualities. Adolescence adds a new story, less stable, very liable to arrest at any stage This makes nearly the whole post-pubic period critical, peculiarly exposed to dangers from without, because it is so plastic and susceptible, and still more so be- cause the growth forces that push youth on toward maturity are so liable to show signs of exhaustion before their work is finished. Hence it follows that length of the growing period is one of the most important factors in development. Lower races often stop short when sexual maturity is achieved." ('03, p. 811.) In this connection, Libby ('08), reporting the result of an experiment in which he had white high school and grammar grade pupils write compositions descriptive of a certain sentimental picture, states that the feeling, meaning and sentiment of the picture were grasped only by pupils older than fourteen years. And Ellison, as reported by Bagley ('09), says that children below the age of thirteen do not have abstract ideas such as would enable them to give good defini- tions. Odum ('10, pp. 36-37) agrees with the general idea repre- sented in the last two quotations from Hall: "Negro children are easily interested, attentive, eager and alert. For the most part they are bright and learn easily. In many cases they ap- pear brighter than white children of the same age. They learn from memory easily and retain little things for some length of time They learn readily to do things by imitation and become comparatively skilful in a short time. .... However, there are many negro children who have an almost total lack of mental perception, whose minds are so dense that they can scarcely learn anything. The percentage of such cases increases with age." This author makes many statements, without, however, giving evidence to substantiate them, to the effect that the negro child, as shown by experi- ments, is brightest — i. e., most able to do and learn simple things — at thirteen years of age, and that he is of greatest ability — i. e., most able to "grasp and hold that which con- fronts the mind" — at eight or nine years of age; that the negro's mental development practically ceases at the age of about thirteen ; that there is an almost entire absence of sexual morality among the great body of negroes, children and adults, due to the predominance of their "physical impulses and pleasure-pain feelings ;" that the more primal emotions, fear, 8 THE PSYCHOLOGY OF THE NEGRO. anger, jealousy, self-exaltation, self -depreciation, sorrow, etc., are especially active in the negro ; that dynamically the negro is volatile, easily responsive to stimuli, guided by present im- pulses, unrestrained — in short, that his life is one of temporary emotion rather than of permanent sentiment. A recent article by Bardin ('13) argues that there must be a connection between racial mental differences and the physical differences between races, since both were evolved together. And since it is becoming increasingly evident that the negro and the white differ mentally, we must therefore suppose that there are corresponding neural differences, as marked, in their way, as are the external physical signs of race, such as skin, hair texture and facial angle. From this posi- tion the writer contends that in attempting to modify the negro's mind while yet keeping him a physical negro we are undertaking the impossible. "It follows, therefore, that pres- ent ideals in regard to the solution of our Negro problem are biologically fallacious, and impossible of attainment. We can never make the Negro like the white man mentally. We can never have a bi-racial state based upon an identity of ideas and political philosophies in both races." ('13, p. 374). This contention is almost identical with that quoted above from Le Bon. One of the most interesting estimates of the intelligence of the negro was made by Francis Galton ('92). In a study of the relative capacity of the white and negro races he divided each race into sixteen defined grades of ability, eight above and eight below its racial average, and considered that the intervals separating the grades were equal throughout. After a survey of eminent men of each race he came to the conclu- sion that the ablest negro ranked two grades below the ablest white. Then, by an application of the "law of deviation from an average," he held that negroes as a race have two degrees of ability less than Europeans. Another way of expressing it is to say that the difference between negroes and whites in intellectual capacity is about one-eighth of the difference between the most eminent man and the lowest idiot. Galton then goes on to point out that the experiences of travelers among native tribes and the prevalence among negroes of feeble intellects furnish confirmation of his estimate. After a survey of the available evidence of racial mental differences, Thomdike ('10, pp. 67-68) sums up as follows: REVIEW OF WORK PREVIOUSLY DONE. tf "From all these facts each student may make his own esti- mate of the original mental differences of races, and learn at least the need of more actual measurements of race differences and of intelligence in interpreting them. My own estimate is that greater differences will be found in the case of the so- called 'higher' traits, such as the capacity to associate and to analyze, thinking with parts or elements, and originality, than in the case of the sensory and sensori-motor traits, but that there will still be very great overlapping. Calling the differ- ence between the original capacity of the lowest congenital idiot and that of the average modern European 100, I should expect the average deviation of one pure race from another in original capacity to be below 10 and above 1, and the differ- ence between the central tendencies of the most and the least gifted races to be below 50 and above 10. I should consider 3 and 25 as reasonable guesses for the two differences." Experimental Studies From these studies and opinions of a non-experimental nature we may turn to those which are based upon quantita- tive investigation. The first attempt at a quantitative study of the negro with which the writer is familiar is that by Bache, published in 1895. This investigator starts with the assump- tion that the more inferior the race, the quicker the reaction time. "That the negro is, in the truest sense, a race inferior to that of the white can be proved by many facts, and among them by the quickness of his automatic movements as com- pared with those of the white." ('95, p. 481). The results of the tests on twelve whites, eleven Indians and ten negroes showed the whites to be the slowest and the Indians to be the quickest with auditory, visual and electrical stimulation. The speed of the negroes was roughly midway between that of the Indians and the whites. The writer explains that the reason the negroes were slower than the Indians was that they were of mixed white and negro blood and had inherited the effects of slavery, while the Indians' mode of life compelled them to rely upon quick movement. This explanation is offered to strengthen the writer's contention that the Indian is a higher race than the negro, and consequently should have a slower re- action time. The study is hardly conclusive; the numbers tested were too small, and the variabilities of the average re- 10 THE PSYCHOLOGY OF THE NEGRO. action times are not given. Certainly the initial assumption was not proved by the tests. Smith made some association and memory tests upon a "typical" negro boy sixteen years of age. The nature of the tests is not given, nor are the results set forth. The author's general conclusion, plainly unwarranted on the basis of the work done, is as follows: "The negro child is psychologically different from the white child. In automatic power he is superior, but in the power of abstraction, of judgment and analysis he is decidedly inferior. This fact must be recognized in the school training." ('96, p. 60). Stetson ('97) gave a memory test to five hundred white and five hundred colored children in the fourth and fifth grades of the schools of Washington, D. C. The average age of the white children was 11 years; the average age of the colored children was 12.57 years. The test was somewhat crude. It consisted in reading to the children four verses of four lines each, explanations of the verses being given and the class re- peating them twice in concert. Later the verses were repro- duced orally to the experimenter by each child, and the repro- ductions were scored as being 100, 75, 50, or 25 per cent, cor- rect. The results showed that the negroes were superior to the whites in memory of three of the verses, while the whites were superior in memory of one. The average score for the four verses was: Whites, 58.09; Colored, 58.27. In other words, there was practically no difference in memory capacity between the two races. But in school studies the average rank of the white children was 74.32, while the average rank of the colored was 64.73. This superiority of the whites in school work led the author to conclude that the negroes were deficient in reasoning power, since the test showed that they were not deficient in memory. This conclusion, of course, is subject to criticism on the ground that a number of other fac- tors may have been responsible for the academic inferiority of the colored children. McDonald studied 91 colored children by means of physical and mental tests. His conclusions are summarized as fol- lows: "Among the boys and girls the per cent, of long heads is much greater after puberty than before. This is also true of white boys but not of white girls. The colored boys are more sensitive to heat and locality after puberty than before. The reverse is true with the white boys, but the colored girls, REVIEW OF WORK PREVIOUSLY DONE. H like the white girls, are less sensitive after puberty Colored children are much more sensitive to heat than white children Colored girls have larger circumference of head at all ages than white girls White children, rela- tively to their height, are longer bodied than colored children. The percentage of long-headedness among colored boys is more than double that of white boys." ('99, pp. 1141-1143). The writer states that from a table based on teachers' estimates of the brightness, dullness and mediocrity of colored children, (number indefinite), it appears that the percentage of bright children, both boys and girls, increases rapidly between the ages of seven and eight, and continues to increase slightly until the age of thirteen. From thirteen to sixteen the per- centage of bright pupils decreases rapidly. It would seem that the results of this work should be verified by other inves- tigators before being accepted. The chances for error in the mental tests used are considerable, and the composition of the various groups tested is not quite clear. The foregoing experimental studies, all of which were made prior to 1900, emphasize the need of a careful technique in the quantitative handling of this question. The opportuni- ties for error are very great, and in inexperienced hands psychological tests and statistical methods may lead to results that are worse than useless. A field as little worked and as inviting as this of the comparative psychology of the white and the negro is likely to attract, and has attracted, investigators who lack the necessary training— just as it has attracted theo- rists who were not adequately grounded in the essentials of their work. One would not be far wrong in saying that all of the experimental work done on the psychology of the negro prior to 1900 is of practically negative value. In summarizing the status of scientific race psychology in 1910, Woodworth writes as follows:* "One thing the psychologist can assert with no fear of error. Starting from the various mental processes which are recognized in his text- books, he can assert that each of these processes is within the capabilities of every group of mankind Statements to the contrary, denying to the savage powers of reasoning, P^rTJi'^p^"'"^ rrif'^^*^,,^^ Woodworth was that by himself, Bruner. Sus^velv ';?th' ,^^^^"^^" ^"d Myers This work had to do almost exl mSrll w 1 \f ""'^i capacities of primitive groups, and the sum- H here ^°°'^'^°^*^ "^^^^« '^ unnecessary to give a further account of 12 THE PSYCHOLOGY OF THE NEGRO. or abstraction, or inhibition, or foresight, can be dismissed at once. If the savage dilf ers in these respects from the civilized man, the difference is one of degree, and consistent with over- lapping of savage and civilized individuals." ('10, pp. 3-4). Woodworth then goes on to discuss the evidence in regard to the several sense capacities. Natives of Brazil, the steppe- dwelling Kalmucks, Papuans, Indians, Filipinos and other races have been tested for visual acuity, and found, on the whole, to have vision superior to that of the average white, but the overlapping between the whites and these races is great. "We may perhaps conclude that eyesight is a function which varies somewhat in efficiency with difference of race, though with much overlapping It did not seem possible, how- ever, to assert anything like a correspondence between eye- sight and the degree of primitiveness or backwardness of a people Even if small differences do exist, it is fairly certain that the wonderful feats of distant vision ascribed to savages are due to practice in interpreting slight indications of familiar objects." ('10, pp. 5-6). In the case of hearing, the tests indicate that whites are superior to primitive peoples. This superiority may be due in part to the fact that the ears of civilized man are better protected from injuries than are those of savages, and that the meatus is kept cleaner by the white man. Then, too, the white is more familiar with the sorts of sound used in the tests than is the savage, and on this account may detect them more readily. The few tests that have been made for keenness of smell show no higher acuity among negroes and Papuans than among Europeans. In ability to discriminate two points on the skin the evidence is conflict- ing ; on the whole there is probably no appreciable superiority in favor of any of the races tested. The experiments which have been made to determine the acuity of the pain sense are largely vitiated by the fact that savages and civilized men have different standards as to what constitutes pain. "On the whole," says Woodworth, "the keenness of the senses seems to be about on a par in the various races of mankind." ('10, p. 7). In reaction time, speed of tapping and suscepti- bility to illusions, the tests seem to indicate that the different races are about equal. In discussing tests for intelligence as opposed to sensory and motor capacity, the writer points out that the form-board is the only test of intelligence which has been used with differ- REVIEW OF WORK PREVIOUSLY DONE. 13 ent races. 'This test was tried on representatives of several races, and considerable differences appeared. As between whites, Indians, Eskimos, Ainus, Filipinos and Singhalese, the average differences were small, and much overlapping occurred. As between these groups, however, and the Igorot and Negrito from the Philippines and a few reputed Pygmies from the Congo, the average differences were great and the overlapping was small. Another rather similar test for intelligence which was tried on some of these groups, gave them the same rela- tive rank. The results of the test agreed closely with the general impression left on the minds of the experimenters by considerable association with the people tested. And, finally, the relative size of the cranium, as indicated, roughly, by the product of its three external dimensions, agreed closely in these groups with their appearance in intelligence and with their standing in the form test. If the results could be taken at their face value, they would indicate differences of intelli- gence between races, giving such groups as the Pygmy and Negrito a low station as compared with most of mankind." ('10, pp. 10-11). One of the important investigations of the mental capacity of the negro is that by Mayo ('13), who studied the school marks of 150 white and 150 colored high school pupils in the schools of New York City. His results can best be given in his own words : "To summarize, the following are the leading results deduced from the data considered: "The median age of white pupils at the time of entering high school in the city of New York is 14 years 6 months ; of colored pupils 15 years 1 month — a difference of 7 months. The average deviation for whites is 9 months ; for colored, 15 months. Twenty-seven per cent, of the whites are as old as the median age of the colored or older. "Colored pupils remain in school a greater length of time than do the whites. For the cases studied, the average time spent in high school for white pupils was 3.8 terms ; for col- ored, 4.5 terms. About 28 per cent, of the whites attain the average time of attendance for the colored. "Considering the entire scholastic record, the median mark of the 150 white pupils is 66 ; of the 150 colored pupils 62 ; a difference of 4 per cent. The average deviation of white pupils is 7 ; of the colored 6.5. Twenty-nine per cent, of the colored pupils reach or surpass the median mark of whites. 14 THE PSYCHOLOGY OF THE NEGRO. "The white pupils have a higher average standing in all subjects The percentage of colored pupils reaching the median mark of the whites in the several subjects is as follows : Modern languages, 33 ; mathematics, 32 ; history, 31 ; the sciences, 29; Latin and Greek, 27; English, 24; the com- mercial subjects, 22; and all subjects together, 29. "The total number of subjects pursued by the white group was 2433; the total number of subjects passed on the first trial was 1855; the percentage of subjects passed being 76. The total number of subjects pursued by the colored group was 2382; the total number of subjects passed on first trial was 1379, the percentage of subjects passed being 58. Inter- preting these figures as a measure of relative scholastic efficiency, the efficiency of colored subjects is 76 per cent, of that of the whites ; that is, the colored pupils are about % as efficient as the whites in the pursuit of high school studies." ('13, pp. 44-45). These results are significant as they stand, and they be- come still more so when it is considered that the colored pupils studied were, as Mayo points out, a more closely selected group than the whites. How much more closely the negroes were selected than the whites is not known. It must also be borne in mind that the colored group was not made up of persons of pure negro blood. The percentage of mulattoes is not stated, but it is probably high. And the presence of mulattoes considerably raises the standard of negro attainment, as will be shown in a later chapter. Another consideration tending to emphasize the racial differences found by Mayo is that the colored pupils with whom he dealt were for the most part either emigrants from the South or the children of emigrants, and that they therefore probably inherited the ability and energy which leads the ambitious negro to seek to better his condition in the North. On the other hand, it is difficult to estimate the white group represented in this study. It con- tained English, Germans, Irish, Italians and Jews in indefinite numbers, but a random selection of whites was carefully ob- served, and the group is probably typical of the white high school population of New York. Phillips ('12), in a study of retardation in the public ele- mentary schools of Philadelphia, found percentages of re- tardation as follows in schools attended entirely by colored pupils: 68.2, 60.6, 67.3, 70.9, 66.3, 72.8, 58.2, 59.3. The per- REVIEW OF WORK PREVIOUSLY DONE. 15 centages of retardation in the respective school districts in which these eight schools were situated were as follows: 41.8, 44.5, 45.1, 45.1, 37.2, 36.0, 36.0, 33.3. In other words, the per- centage of retardation in the colored schools ranged from 72.8 to 58.2, while the percentage of retardation in the districts which contained these schools ranged from 45.1 to 33.3. The average percentage of retardation for the city as a whole was 40.3. Each of the colored schools had a greater percentage of retardation than any of the white schools, even those com- posed almost entirely of foreigners, and in those schools at- tended by both white and colored pupils the percentage of re- tardation on the whole varied directly with the percentage of colored pupils in attendance. The writer concludes by saying : "It is a question whether the course of study is suited to the negroes, as the educational results are so far behind those in the other schools, and it is very doubtful whether even a liberal interpretation of the course of study would meet the educational necessities of this group." ('12, p. 90). In a later article ('14), Phillips reports the results of an attempt to ascertain the causes of this retardation by means of an application of the Binet tests to white and colored chil- dren of the same chronological age and home conditions, the tests being made in all cases by the same individual. "Forty colored girls and 46 colored boys, totaling 86, were tested by the Binet scale ; 75 white girls and 62 white boys, totaUng 137, were likewise tested. The home of each of these 223 pupils was visited and the home conditions noted, as Excellent, Good, Fair and Poor. In so rating the home, the material (money) , intellectual, and moral elements were noted in making up the rating. In the following comparison only the white children of excellent home conditions are compared with the colored children of excellent home conditions ; the white of good home conditions with the colored of good home conditions, and so on. This method of procedure, of course, necessitated the elimina- tion of quite a number of those tested, so that our final com- parison was made on 29 each of colored boys and girls re- spectively." ('14, p. 191). The results are stated as follows : " . . . . we see that of those tested 37.9 per cent, of the white boys were retarded, while 65.5 per cent, of the colored boys were retarded; that 46.4 per cent, of the white girls were retarded and 71.4 per cent, of the colored girls were retarded ; and that 42.1 per cent. 16 THE PSYCHOLOGY OF THE NEGRO. of the white boys and girls combined were retarded, and that 68.4 per cent, of the colored boys and girls combined were re- tarded. This makes the colored boys retarded 17.4 (27.6)* per cent, more than the white boys ; the colored girls retarded 25 per cent, more than the white girls; while the total rate of retardation of both groups is 26.3 per cent. " we find that 31 per cent, of the colored girls (boys)t are accelerated while 62 per cent, of the white girls (boys)t are accelerated; that 28.5 of the colored boys (girls) f are accelerated and 53.5 per cent, of the white boys (girls) ;t and that 29.8 per cent, of colored boys and girls are accelerated and 57.8 per cent, of both white girls and boys. This makes a difference in the acceleration between the two races of 31 per cent, in favor of the white boys, 25 per cent, in favor of the white girls, 28 per cent, in favor of the white pupils with boys and girls combined. "This would seem to corroborate the findings in the case of pedagogic retardation. We see in every group, considering the retardation from pedagogical or psychological viewpoints, that the colored pupils are retarded from 20 to 30 per cent, more than the white pupils, and that the white pupils are always greatly above them in acceleration." ('14, pp. 191-195). It is interesting to note that the total number of pupils tested, including those eliminated on account of not having comparable home conditions, gives practically the same com- parative result as was obtained from the picked group. The percentage of retarded pupils in the total colored group was 54.6; the percentage of retarded pupils in the total white group was 24. The percentage of accelerated pupils in the total colored group was 6.9; in the total white group it was 20.4. The author concludes as follows: "In applying the Binet tests to colored children the following facts of interest were fortunately thrust upon our attention. In the first place the colored pupils as a class were good in the memory tests and poor in those requiring judgment. They were generally *This figure should evidently be 27.6 instead of 17.4. It represents the difference between the percentages of retardation of white and col- ored boys, which are 65.5 and 37.9, as quoted, and also as given in the tables which accompany the article. fPhillips seems to have made an error here. In order to make the i?"^j^4^^^®® with the tables from which they are taken, and also with the differences in acceleration as quoted, the words "boys" and "girls" should be interchanged as indicated by the brackets. REVIEW OF WORK PREVIOUSLY DONE. 17 slower in response. The testing of the colored children took a much longer time than the white. Their reaction time was greater, they were less animated. It is significant to note that the younger white children were more advanced than the colored children of the same age. This is in contradiction to the generally accepted fact that colored children are quicker when young. "If the Binet tests are at all a gauge of mentality it must follow that there is a difference in mentality between the col- ored and the white children, and this raises the question: Should the two groups be instructed under the same cur- riculum?" ('14, p. 196). The first application of the Binet tests to whites and ne- groes was made by Strong, and was reported by her ('13) and by Morse ('14). Two hundred and twenty-five white and 125 colored children were tested in the schools of Columbia, S. C. The percentage of children testing more than one year below age was 10.2 for the white and 29.4 for the colored. The percentage testing more than one year above age was 5.3 for the white and 0.8 for the colored. The largest group of white children was that testing at age; the largest group of colored children was that testing one year below age. An attempt was made to divide the white children into "city children" and "mill children," in order to arrive at a conclusion as to how far inferiority in the tests was due to poor environment. The environment of the mill children is of considerably lower grade than that of the city children, and is not markedly different from that of the negroes. When this division was made the results showed that 6 per cent, of the city children were more than one year below age while 18 per cent, of the mill children were more than one year back- ward. On the face of it, then, this would indicate that the comparatively poor showing made by the negroes was in large measure due to poor home conditions. But in fact it leaves the question still open. For while mill children may have ad- verse environment, they may also have poor native capacity due to their poor heredity. Their unfavorable surroundings may be the product of a lack of inheritable capacity in their parents. This criticism applies to all attempts to determine the in- fluence of environment upon people whose heredity is not known to be alike. And yet such efforts are made with 18 THE PSYCHOLOGY OF THE NEGRO. seemingly increasing frequency. It should be remembered that poor achievement, in psychological tests or in any other activity, may be accompanied by inferior social condition and yet not be the product of that condition. The achievement and the condition may both result from one and same cause, lack of native capacity. In such matters as are sought by all, such as the ordinary facilities of life and intercourse accord- ing to the prevailing standards, it is very probable that native capacity determines the relative attainment of men within any large and homogeneous unit of mankind. And it is very probable that the presence of poor home environment among any considerable group of our population is due to a compara- tive absence of ability in that group as a whole. This ab- sence is inherited by the offspring of that group. Errors due to a neglect of this consideration need to be especially guarded against in psychological investigations which aim to study environmental influence. Strong divided the negroes tested by her into classes on the basis of degree of skin pigmentation. She says: "The children were divided into three groups according to color. This classification was not a scientific one, and the statement of results may be entirely worthless. There were 34 dark children, 35 medium in color, and 43 light colored in this classification, 122 in all. Of the dark colored, 14.4 per cent, tested below age, 76.7 per cent, tested at age, and 8.8 above age. Of the next group, somewhat lighter in color, 31.1 per cent, tested below age, 62.2 at age, and 6.6 per cent, above age. Of the lightest group 44.2 tested below age, 44.2 at age, and 11.6 above age. The darkest children are more nearly normal, the lightest show the greatest variation, both above and below normal." ('13, p. 506). Morse comments upon the tests as follows: "In general it may be said that the colored children excel in rote memory, e.g., in counting, repeating digits (but not one was able to re- peat 26 syllables), naming words, making rhymes and in time orientation. They are inferior to the whites, however, in aesthetic judgment, observation, reasoning, motor control, logical memory, use of words, resistance to suggestion and in orientation or adjustment to the institutions and complexities of civilized society." ('14, p. 78). With reference to peda- gogical retardation, he writes : " according to the Binet scale, a larger number of white children are in a school grade REVIEW OF WORK PREVIOUSLY DONE. 19 below their mental ability than above, whereas the reverse is true of the colored children." ('14, p. 78). In reviewing this study, Bruner says: "The tables show another interesting point on which the author makes no com- ment. At the ages of six, seven and eight just about twice as many negro children as white rate below age, whereas for the ages of ten, eleven and twelve the superiority of the whites over the negroes is but slight. This suggests that the rate of maturing may be more rapid with the negro children, so as to make them older, mentally, at the age of twelve than white children of the same age." ('14, p. 385). A study of the learning capacity of whites and negroes was made by Baldwin ('13). A somewhat elaborate substi- tution test was used five minutes a day for sixteen days with 37 white and 30 colored delinquent adolescent girls. Their ages ranged from 13 to 21 years. "Fourteen other negro girls were too feeble mentally to perform the tests after the initial instructions although they worked assiduously for the period of three weeks, and three white girls failed to do 50 per cent, of the work correctly." ('13, p. 317). From the summary of the results: "In this type of learning it is found that: 1. Comparing the amount of work done by the thirty- seven white girls with the work done by the thirty negroes who accomplished more than 50 per cent, of correct results, it is evident that the negroes are decidedly inferior. The white girls made 72.3 substitutions as a general average, the negroes 55.8. The negroes accomplished 62.4 per cent, as much work as the white girls and made 245.3 per cent, as many errors. Practically all the superior negroes in the school were included in the test. "2. The learning capacity of delinquent negro girls differs quantitatively and qualitatively from that of the white girls, and the educational corollary follows that different methods of instruction and training are required for the negro girls than for the white girls." ('13, pp. 331-332) . On the whole the inferiority of the negroes was about the same in both absolute amount of work done and in learning capacity. The negroes as compared with the whites were slow to warm up, quick to lose interest, difficult to stimulate except through flattery, irregular, moody, vacillating in atten- tion, inaccurate, envious of each other's progress, given to 20 THE PSYCHOLOGY OF THE NEGRO. mumbling, grumbling, humming, saying funny things while at work. A study of this kind is difficult to evaluate in its bearing upon normal subjects. The girls were all committed to an institution as delinquents, and their mental ability was proba- bly considerably below the average. Whether both races were equally below the average is not known. The fact that fourteen negro girls could not work the test at all, while only three white girls failed to complete as much as 50 per cent, of the work, would indicate that the negroes were farther below the average racial ability than were the whites; but this is not conclusive. That the test failed to enlist the in- terest of the colored girls indicates that their true learning ability was not measured. But this, too, is of doubtful sig- nificance, since the very fact that the negroes were not inter- ested as were the whites possibly points to a deficiency in the colored group. On the whole, it seems safe to say that the test is probably indicative of a true racial difference, though the amount of such difference is left uncertain. Since the tests to be described in the following chapters were made, there has appeared a preliminary report by Pyle ('15) of an extensive investigation of the relative mental capacity of whites and negroes. This investigator tested 408 colored pupils, from eight to sixteen years of age, inclusive, in the public schools of Columbia, Mexico and Moberly, three towns in Missouri, using a number of standard tests, and com- pared the results thus obtained with the norms which he had previously ascertained for white children in the same tests. As the report which has so far appeared is only preliminary, it is impossible to go into detail concerning the work done, but the author's words may be quoted to indicate the main out- lines of this valuable research: "The following are the conclusions to which the work so far points In general the marks indicating the men- tal ability of the negro are about two-thirds those of the whites. The negro girls approach the white girls in ability a little more closely than the negro boys approach white boys. Negro boys and girls are farther apart in ability than are white boys and girls. In both races the girls are superior, if the average performance is taken as the basis of compari- son. With increasing age, there is a tendency for the differ- ence between whites and negroes to become less. This ten- REVIEW OF WORK PREVIOUSLY DONE. 21 dency is more marked with boys than with girls. About one- fifth of the negroes are equal or superior to the average of the whites, while three-fourths of the whites are equal or superior to the average of the negroes. In the same school grade, the negroes are several months older than the whites. Negro girls have the best permanent memory for ideas in the eleventh year. The same is true of white girls. Negro boys have the best permanent memory for ideas in the twelfth year and white boys in the thirteenth year. In rote memory the negroes have a much better memory for concrete than for abstract words, but are greatly inferior to whites in both. "If, for purpose of comparison, the negroes are sep- arated into two groups according to social position, it is found that the negro boys of better social class have about four-fifths of the ability of white boys. The negro girls of better social position have an ability which is about three- fourths that of white girls. Difference in social position has less effect on negro girls than on negro boys. The difference in social position has most effect on tests requiring quickness in learning, quickness in controlled association, in immediate and permanent logical memory and in constructive imagina- tion as measured by the Ebbinghaus test. With negro chil- dren of the better social class the tendency to approach the florm of white children is more marked. In the substitution, controlled association and Ebbinghaus tests, the negroes are less than half as good as whites. In free association and the ink-blot tests they are nearly as good. In quickness of perception and discrimination and in reaction, the negroes equal or excel the whites. "At all ages, the physical development both in muscular strength and muscular speed is nearly the same for negro boys and white boys. The same is true for negro girls and white girls until the age of ten. After ten, negro girls are stronger than white girls, but white girls are faster. The negro girls are stronger probably because they do more mus- cular work than white girls do. Muscular speed seems to be little affected by conditions of life while muscular strength is much affected by them. "Perhaps the most important question that arises in con- nection with the results of these mental tests is this : How far is ability to pass them dependent upon environmental condi- tions? Our tests show certain specific differences between 22 THE PSYCHOLOGY OF THE NEGRO. negroes and whites. What these differences would have been had the negroes been subject to the same environmental in- fluences as had the whites, it is difficult to say. The results obtained by separating the negroes into two social groups would lead one to think that the conditions of life under which the negroes live might account for the lower mentality of the negroes. On the other hand, it may be that the negroes living under the better social conditions are of better stock. They may have more white blood in them." ('15, pp. 357- 360). Neurological Studies It is assumed everywhere as a matter of course that men- tal differences imply neural differences. If there are mental differences between two races, then we may expect to find differences in the nature of the brain structure of the two races, and vice versa. The comment is frequently made that apparent differences in the size or shape of the brains of given racial groups must indicate corresponding psychical divergencies. Such comment has been quoted in some of the foregoing abstracts, notably in those from Boas, Le Bon and Tylor. If there is a racial, or individual, inheritance of mind there must be an inheritance of appropriate anatomy and physiology. A few recent investigators have attacked this problem of racial neural differences. In Baltimore, Bean studied the brains of 103 negroes and 49 Caucasians. He concludes that, "Not only is the anterior association center smaller in the Negro than in the Caucasian, but the whole frontal lobe of the Negro is smaller." ('06, p. 374). The negro stands in an intermediate position between man and ourang in the rela- tive size of his frontal to his parietal and occipital lobes. Bean then goes on to point out that this conclusion is in accord with well-known traits of the negro. He states, citing Flechsig as authority, that the anterior association center, which is comparatively small in the negro, is intimately con- nected with ideas regarding personality ; the relations of self, subjectively and objectively; the capacity for ethical and aesthetic judgment; self-control, especially in such matters as sexual excitement, anger or vexation; will power. The pos- terior association center, on the other hand, which is com- paratively large in the negro, is more intimately connected REVIEW OF WORK PREVIOUSLY DONE. 23 with the special senses; it is objective and concrete, while the anterior center is subjective and abstract in the mental processes which its operation accompanies. "The relative differences found in the association centers of the two races is suggestive in relation to the known characteristics of the two, in view of Flechsig's work. The Caucasian i« subjective, the Negro objective. The Caucasian .... is dominant and domineering, and possessed primarily with determination, will- power, self-control, self-government, and all the attributes of the subjective self, with a high development of the ethical and aesthetic faculties. The Negro is in direct contrast by reason of a certain lack of these powers, and a great development of the objective qualities. The negro is primarily affectionate, immensely emotional, then sensual and under stimulation passionate. There is love of ostentation, of outward show, of approbation ; there is love of music, and capacity for melodious articulation; there is undeveloped artistic power and taste — Negroes make good artisans, handicraftsmen — and there is instability of character incident to lack of self-control, especially in connection with the sexual relations ; and there is lack of orientation, or recognition of position and condition of self and environment, evidenced by a peculiar bumptiousness, so-called, that is particularly noticeable. One would naturally expect some such character for the Negro, because the whole posterior part of the brain is large, and the whole anterior portion small, this being especially true in regard to the anterior and posterior association centers." ('06, pp. 378- 379). Further, Bean reports that he found the ratio of the corpus callosum to the total brain weight to be greater in the Cau- casian than in the negro, the anterior end of the corpus cal- losum in the whites being relatively large when compared with its posterior end. In weight, the 51 negro brains, male, averaged 1292 grams, while the 37 white male brains averaged 1341 grams; the 28 female negro brains averaged 1108 grams, while the 9 white female brains averaged 1103 grams. The negroes were of a higher class than the whites, however, and mulattoes were included among them. These conclusions of Bean's are very interesting, but they seem to need further confirmation before they can be ac- cepted as final. Following Bean, Mall ('09) found no such 24 THE PSYCHOLOGY OF THE NEGRO. definite racial differences. He states that the brain weight of eminent men is 100 grams above that of men in general, and that the average white man has a brain 100 grams heavier than that of the average negro. But the frontal lobe as com- pared with the rest of the brain has the same relative weight in both negroes and whites, male and female. And the con- figuration of negro and white brains is the same. " with the present crude methods the statement that the negro brain approaches the foetal or simian brain more than does the white is entirely unwarranted." ('09, p. 20). "It certainly would be important if it could be shown that the complexity of the gyri and sulci of the brain varied with the intelligence of the individual, that of the genius being most complex, but the facts do not bear this out, and such statements are only misleading brains rich in gyri and sulci, of the Gauss type, are by no means rare in the American negro." ('09, p. 24). This investigator reviews the previous work done in this field, and comes to the final conclusion that there is no valid evidence to show significant brain differences from the point of view of race, sex or genius. Karl Pearson ('07), after a study of the heads of 1000 Cambridge graduates and 5000 school children, states that his results corroborate the conclusions of previous articles, and sets forth his findings as follows: "The average correlation between head length or head breadth and intelligence is .11. .... no sensible modification is made in this result if allow- ance be made for either weight or stature." But "some 44 per cent, of very able men have heads smaller than the aver- age slow man and some 44 per cent, of slow men heads larger than the average specially able man. This order of numerical relationship holds for the whole range of the characters dealt with, and in view of it we see how idle it is to assert that head measurements can be of any service in the prediction of in- telligence Differences in size of head will not account for at most 1/12, and probably not as much as 1/20, of the observed differences of capacity whether between adults or be- tween children." ('07, p. 121). Hrdlicka ('98) investigated the physical differences be- tween 1100 white and 300 negro children, and found that the negro's forehead is narrower but not lower than that of whites, and that the negro's head is unusually long and nar- REVIEW OF WORK PREVIOUSLY DONE. 25 row, while the heads of whites are of all shapes. He says the negroes' ears are smaller, their arms, hands and feet longer, and their chest somewhat deeper than is the case among white children. The weight of white children at all ages is somewhat greater than that of negroes, but the negroes at all ages and of both sexes are three or four pounds stronger with each hand. Hrdlicka also finds, as did Le Bon in the abstract previ- ously given, that the variation among the whites is greater than among the negroes. "The white children show more diversity, the negro children more uniformity in their normal physical characters. This fact becomes gradually more marked as we advance with the age of the children." ('98, p. 476). it is impossible to make an adequate summary of the views set forth in this chapter ; the abstracts given are them- selves summaries. But it is clear that by far the greater num- ber of writers who have dealt with the problem of the rela- tive mental ability of the white and the negro take the view that the negro is inferior. This is particularly true of those investigators who have used quantitative methods. The negro has not shown the same capacity as the white when put to the test of psychological or educational experiment, and the racial differences revealed have been considerable. In the higher mental processes that go to make up the capacities necessary to a successful conduct of civilized life, the negro seems to fall short to a far greater degree than in the ele- mentary traits which man has in common with the lower ani- mals. In sense capacity, in instinct, in motor ability, there is no evidence that he inferior to the white man. It is in such matters as reasoning ability, the power to perceive relations, to exercise creative imagination, to subordinate a present pas- sion to a distant end, that the weight of evidence and opinion indicates his relative deficiency. With regard to the comparative equality of white and negro children up to the age of adolescence and the then en- suing superiority of the whites, the evidence is not at all clear. The theories as to the significance of adolescence upon which this view is based, have themselves undergone restricting modification in very recent years, and the whole matter is at present unsettled. There may be greater differences between 26 THE PSYCHOLOGY OF THE NEGRO. white and negro children either before or after adolescence, so far as is known. Or adolescence may have no appreciable bearing upon ascertainable racial differences. The question of differences in brain structure between the two races is likewise subject to controversy. It seems to be indubitable that in brain size there is a difference, but the in- ternal structure of the brain, which is far more significant for intelligence than size, is as yet a subject for debate rather than for evidence in so far as it concerns differences between whites and negroes. The abstracts given are believed to include all of the experimental studies and a fair sampling of the better studies which are not experimental. It is probably true that there are more people who believe in racial mental equality than the reviews would indicate; equality is taken for granted, as in the greater part of our school system and in our political life ; it is those who believe in racial inequality who consider their views novel enough for publication. It may be said that the main conclusion one may draw from a study of the litera- ture bearing upon the mental side of our race question is that we have taken a step toward its solution, but that the problem is still a problem. The evidence with regard to the relative ability of pure negroes and mulattoes will be discussed in Chapter IV, as will that bearing upon racial variability. CHAPTER II THE SUBJECTS AND THE TESTS The Subjects The experiments to be described herein were made in De- cember, 1914, upon pupils in the schools of Richmond, Fred- ericksburg and Newport News, Virginia. According to the Census for 1910 ('10), Richmond has a population of 127,628, of whom 36 per cent, are negroes; Fredericksburg has a population of 5874, of whom 25 per cent, are negroes; and Newport News has a population of 20,205, of whom 36 per cent, are negroes. In the State as a whole, 32.6 per cent, of the population are negroes. The white inhabitants of these cities are to a very large extent native born and of native parentage ; in the state as a whole 95.4 per cent, of the white population is native born and of native parentage, and in these cities the percentage does not considerably differ from that of the state. The Census shows that in Richmond 1.2 per cent, of the native white population ten years of age and over are illiterate, while 19.6 per cent, of the negroes are illiterate. In Fred- ericksburg the corresponding percentage of white illiteracy is 1.5; that of negroes is 20. In Newport News the per- centages are : whites, .6 ; negroes, 12. In the state as a whole, 8 per cent, of the white population and 30 per cent, of the col- ored population are illiterate. It is thus evident that the cities have a smaller percentage of illiteracy, both white and colored, than the state, and that a much greater proportion of the negroes than of the whites is illiterate. The percentages of the population 6-14 years of age that attend school are as follows, according to the Census: Rich- mond — native whites, 79.2 per cent.; negroes, 65.2 per cent. Fredericksburg — whites, 76 per cent.; negroes, 64 per cent. Newport News — whites, 76; negroes, 69 per cent. In the 27 28 THE PSYCHOLOGY OF THE NEGRO. state as a whole 73.2 per cent, of the white and 58.7 per cent, of the colored population 6-14 years of age attend school. It thus appears that a somewhat larger percentage of whites than of negroes attend school, and that this difference between white and colored school attendance is only slightly less in the cities than in the state at large. It should be noted that there is no compulsory education law in any of the cities mentioned, and that only an inconsiderable fraction of the population of the state attend school under such a law. So far as these figures show, the white and colored popu- lations of the cities in which the tests were made do not differ significantly from the general white and colored populations of Virginia. The negroes, in the cities and in the state at large, attend school less than do the whites in proportion to their numbers and are considerably more illiterate. This last consideration, that the whites attend school in larger proportion than the negroes, and that the negroes are more illiterate, indicates that on the whole the negroes who do attend school are a more closely selected group than are the whites who attend school. It is a rarer occurrence for a negro to become educated. The school selects for its opera- tions a more circumscribed group of negro than of white chil- dren. The Richmond School Report for 1912-'13 ('14) strengthens this conclusion. The figures in the report show that of the total white school enrollment, 10.53 per cent, are in the high school ; of the total colored school enrollment, 4.97 per cent, are in the high school. That is, that out of an equal number of pupils from each race, there are twice as many white as colored of high school grade. If we compare the percentage of the total school population, white and colored, enrolled in elementary and in high schools, we arrive at the same result. Of the white and colored school populations, there are 54 and 55 per cent., respectively, en- roiled in the elementary schools; but there are 7.5 per cent, of the white and only 3.1 per cent, of the colored school popula- tion enrolled in the high schools. The same percentage of the white and of the colored school populations is enrolled in ele- mentary schools, but there is twice as great a percentage of the white as of the colored enrolled in high schools. Of the colored population as a whole, a smaller proportion is in school as compared with the whites ; and of those in school, a smaller THE SUBJECTS AND THE TESTS. 29 proportion is in high school. It is a rarer thing for a colored than for a white child to attend school at all; it is a much rarer thing for a colored school child to be in high school. Colored high school pupils are the "chosen few" of their race to a greater extent than are white high school pupils. The nature of the selective factors which thus act more intensively upon negroes than upon whites as the school grades advance is difficult to ascertain. Among the causes of elimina- tion from school are such matters as poor health, poverty, mental or moral deficiency, lack of ambition or energy. It would seem that these forces, in the long run, must select for survival in the school system those who by reason of their own capacity, as represented in themselves and in the inherit- able traits which prompt and enable their parents to send them to school, are best fitted to make progress in academic work. It would seem that the school must select as well as train those who have greatest ability and who thus profit most by school attendance. That this is true in the case of the high school's selection of negro pupils is especially indicated by the following considerations. The figures for retardation in the schools of Richmond are as follows: In the white elementary schools, 51.8 per cent, of the pupils are above the normal age for their grade; in the colored elementary schools, 75.0 per cent, of the pupils are above the normal age for their grade. While in the white high school 52.1 per cent, of the pupils are above the normal age for their grade, and in the colored high school, 55.8 per cent, of the pupils are above normal age. The colored elemen- tary pupils are 23.2 per cent, more retarded than are the white elementary pupils; the colored high school pupils are only 3.7 per cent, more retarded than are the white pupils of their grade. It is evident that the pupils who do not accomplish what is expected of their age drop out of the negro schools before high school is reached to a greater extent than out of the white schools. Indeed, while it appears from the figures that the colored pupils who are eliminated from the school sys- tem are those who do not perform the work of their age, it does not appear that this is true of the whites. It would almost seem that whereas the negro goes to high school by reason of his ability and determination, the white goes on account of some other incentive, such as, perhaps, social pressure or the custom of his class. It is not unreasonable to con- 30 THE PSYCHOLOGY OF THE NEGRO. elude on the basis of the figures for retardation that one of the selective factors which operate to a greater extent upon colored than upon white pupils is inabihty to do the school work expected of their age. In this connection it should be re- called that Phillips, in a study reviewed in Chapter I, found the relative percentages of white and colored retardation in the elementary schools of Philadelphia to be about the same as those found here ; and that his investigations with the Binet tests showed that the greater pedagogical retardation of the colored pupils was paralleled by their greater psychological retardation. When we divide the elementary school into primary and grammar grades, we find that 35 per cent, of the white ele- mentary pupils are in the grammar grades, while only 21 per cent, of the colored elementary pupils are in these grades. This is in confirmation of our conclusion that the higher grades have a larger percentage of the white than of the colored children. But when we divide the high school into two parts, consisting of the first and second and the third and fourth years, respectively, we find a different situation. Twenty- five per cent, of the white high school pupils are in the third and fourth years combined, while 28 per cent, of the colored high school pupils are in the third and fourth years com- bined. Here we have a larger proportion of colored than of white pupils in the upper years. It is interesting to note that Mayo, in a quotation previously given, found the same situation in the high schools of New York. The colored pupils studied by him remained in school longer than did the white pupils. It seems that after the high school is reached, selec- tive factors eliminate a larger percentage of white pupils than of colored. This tends to corroborate the view expressed above, that the negro who enters high school does so because of his ability and determination, whereas the white high school pupil often enters by reason of social pressure, custom, or the tradition of his race. There is a marked difference between the work of the high school and that of the elementary school. One of the problems of modern education is to lessen the gap between the two. Within either school the work of a given grade is not much more difficult than the work of the grade below it. But the first year of the high school is much more difficult than the last year of the elementary school. This increased diffi- THE SUBJECTS AND THE TESTS. 31 culty causes a large number of the less capable pupils to drop out of high school at the end of the first or second year. A larger proportion of white than of colored pupils so drop out because a greater proportion of white than of colored entered without serious purpose or the requisite ability. The negroes who enter high school are a more closely selected group, and they therefore more nearly finish the course. The figures show that the percentage of retardation in the colored high school is the same for the first two years and for the last two years ; in the white high school there is a considerably greater percentage of retardation in the first two years than in the last two. Another fact which bears upon the relative action of selec- tive factors upon white and colored children in the public schools is that a much larger proportion of white than of colored pupils of high school grade are not enrolled in the pub- lic schools at all, but attend private institutions. This is par- ticularly true in Richmond; and on the whole the pupils who attend private schools are of better social standing, and therefore, on the whole, probably of greater ability, than the average of the school population. The public high school thus loses a number of pupils of ability, and this loss is not felt by the colored high school as it is by the white. Taken all together, the facts brought out show that the colored child in the schools of Richmond, in the upper grades and especially in the high school, is much more closely selected by reason of his ability than is the white child. We should therefore expect the colored pupils of advanced grade to attain a higher score in psychological tests than those of lower grade, when compared with white pupils. This we shall find to be true. And we shall also find, in marked confirmation of the present contention, that colored high school pupils excel col- ored elementary pupils to a much greater extent than white high school pupils excel those of lower grades. Phillips, Strong and Pyle, (see Chapter I), found the mental difference be- tween whites and negroes to become less as the grades ad- vance. The explanation of their findings is probably to be found in this matter of selection. It seems likely that if one could test a random and not an educational selection of whites and negroes, he would not find the difference between the races to decrease with age. And it appears to be certain that racial mental differences discovered by means of tests upon 32 THE PSYCHOLOGY OF THE NEGRO. school children are in reality smaller than the actual differ- ences between the races. The detailed figures for the school systems of Fredericks- burg and Newport News are not available, but there is no reason for believing they would point to a different conclu- sion from that found in Richmond. The Richmond figures are much more reliable than those of the other two cities would be since they are based upon a far greater number of pupils. It may be said that the Richmond figures are derived from the records of 12,018 white and 6184 colored children. The number of pupils tested in these experiments was as follows: In Richmond, 269 white and 319 colored; in Fred- ericksburg, 84 white and 63 colored; in Newport News, 133 white and 39 colored. In all, there were 486 white and 421 colored pupils — a total of 907. The Richmond pupils were in the three years* of the grammar school and the four years of the high school; in the grammar grades there were 149 white and 175 colored ; in the high school there were 120 white and 144 colored. The Fredericksburg pupils were: white, 36 in the 6A and 7A grades and 48 in the high school; colored, 28 in the 6A and 7A grades and 35 in the high school. In Newport News all of the pupils tested were in the grammar grades: 133 white, in grades 6B, 7A and 7B; and 39 colored in grades 6A and 7A. Throughout the study main reliance will be placed upon the results obtained from Richmond, on account of the greater number of pupils tested there ; the re- sults from Fredericksburg and Nevv^port News will be used as corroborative. It may be remarked that in all ways the only difference between the results from the three cities is that those obtained from Fredericksburg and Newport News emphasize the racial differences found to a somewhat greater degree than do those from Richmond. The high schools tested in Richmond were the John Mar- shall, white, and the Armstrong, colored. These are the only high schools in Richmond. In the John Marshall High School *The elementary school in Virginia covers only seven years. The first four years are the primary, the last three are the grammar. The gram- mar grades are 5A, 5B, 6A, 6B, 7A, 7B. In the high school there are four years, and the grades are lA, IB, 2A, 2B, 3A, 3B, 4A, 4B. "A" means the first half of a year; "B" means the second half of a year. THE SUBJECTS AND THE TESTS, 33 there are* 1394 pupils, 535 boys and 859 girls; in the Arm- strong High School there are 76 boys and 257 girls, a total of 333. The elementary schools tested were the Madison, white, and the George Mason, colored. These were chosen because in the opinion of the school authorities their pupils were fairly typical of the average white and colored populations of the city. The Madison School has 871 pupils, 398 boys and 473 girls; the George Mason School has 849 pupils, 360 boys and 489 girls. All four schools are parts of the public school system. In Fredericksburg, the white pupils tested were in attend- ance upon the one public school building for whites in the city. The colored elementary pupils were in the colored public school. But the colored high school pupils were in a private school conducted by colored people, since the city does not maintain a colored high school. This private school conforms, closely in all essential respects to the requirements for pubUc high schools, and its pupils may be fairly compared with those in the public schools. In Newport News, the white pupils were in the John W. Daniel School and the colored pupils were in the John Mar- shall School. These schools were typical of the schools for the two races, in the opinion of the school authorities. No high school pupils were tested in Newport News, since there is no colored high school. In selecting the pupils from the various schools for the tests, in some instances the only grade of a given degree of advancement in the school was tested; in other instances, where choice had to be made among several grades of a given degree of advancement, the school authorities selected a grade of average ability ; in still other instances, where it was neces- sary to test only part of a given grade, the selection of pupils was made by taking them in alphabetical order from the roll. By these means it is believed that the selection of pupils for the tests was made a fair one in all of the schools. That the schools themselves were comparable as between the two races there is no valid reason to doubt. All of them pursued the same general course of study ; within a given city all were parts of the same system, with the exception of the colored high school in Fredericksburg. The teachers and *These are the figures for 1912-'13. The present figures are not available, but they do not differ significantly from those given. 34 THE PSYCHOLOGY OF THE NEGRO. principals of the colored schools were colored in Fredericks- burg and Newport News. In Richmond, the colored elemen- tary school had colored teachers and a white supervising principal ; the colored high school had white teachers through- out. No difference could be perceived in the attitude of the two races toward the tests: both white and colored seemed to enjoy the work rather than the reverse, and both worked with vigor. The number of pupils tested is set forth in detail in Tables 1 and 2. In the treatment of results, no account is taken of a record from only one pupil, and the two lowest and the two highest ages are disregarded on account of the small number of pupils in them. TABLE I. Number of Subjects Tested — Classified by Age and Sex Richmond. Ages 9 10 11 12 13 14 15 16 17 18 19 20 21 Totals White Boys. 1 1 12 15 20 31 17 11 18 5 131 White Girls 11 18 21 27 18 16 20 5 2 .... 138 Col. Boys 1 6 14 19 18 16 14 9 6 4 .... 107 Col. Girls 3 9 22 34 42 27 33 23 15 3 .. 1 212 Fredericksburg. White Boys 3 2 7 11 7 6 6 3 45 White Girls 1 6 6 7 6 6 5 2 39 Col. Boys 1 3 2 1 3 5 1 2 4.... 22 Col. Girls 1 .. 9 10 3 2 10 2 2 1 1 63 Newport News. White Boys 3 15 20 12 5 55 White Girls 5 18 25 23 4 2 77 Col. Boys 1 2.. 4 1 8 Col. Girls 7 6 12 3 3 31 Grades White . Colored TABLE II. Number of Subjects Tested — Classified by Grades Richmond. 5A 5B 6A 6B 7A 7B lA 2A 2B 3A 4A Totals 38 35 19 21 18 18 30 29 .. 35 26 269 ....42 38 35 22 19 19 42 31 21 25 25 319 White . Colored White . Colored 20 21 Fredericksburg. 16 7 14 12 10 9 Newport News. .. 59 40 34 23 .. 16 .. 11 11 8 8 84 63 133 39 THE SUBJECTS AND THE TESTS. 35 A fact should be noted in comparing the results from the white and the colored high schools of Richmond. The white pupils in each of the four years tested were in the first, or "A", half of the year's work. The colored pupils, however, on account of the way they are classified in the school, were in both the first and the second, the "A" and the "B", halves of each of the four years work. To state it in another way, the white high school pupils were in grades lA, 2A, 3A and 4A; while the colored high school pupils with whom they are compared were in grades lA and IB, 2A and 2B, 3A and 3B, and 4 A and 4B. Thus the colored pupils were farther ad- vanced academically than were the white pupils in each of the four years. For the sake of simplicity the high school grades for both white and colored are put down as "A" grades in the tables. But this is somewhat unfair to the white pupils, since the colored were in both "A" and "B" grades in about equal numbers. The ages of the grades tested are shown in Table 3, and the difference in ages between the two races is shown in Table 4. The colored pupils are older, grade for grade. In Richmond, the difference is .36 of a year, in Fredericksburg it is 1.02 years, and in Newport News it is 1.7 years. The difference as shown for Richmond should be reduced somewhat on account of the fact that the colored high school pupils are in reality more advanced in grade than the tables would indi- cate, as was pointed out in the preceding paragraph. Stetson, TABLE III. Ages of the Grades Tested Richmond. Grades 5A 5B 6A 6B 7A 7B lA 2A 2B 3A 4A White Av A.D Col. Av A.D . ... 12.4 , ... 1.1 . ... 12.3 , . . . 1.0 12.5 13.4 13.2 13.3 1.1 1.4 .9 1.0 12.8 12.7 13.5 14.0 .9 .8 1.0 .9 Fredericksburg. 13.6 .9 14.5 .8 14.4 .9 14.9 .7 15.4 1.1 16.2 .8 16.5 1.0 16.5 17.0 .8 .5 16.8 17.6 .7 1.1 White Av A.D Col. Av A.D .. 12.5 .. 13.6 .9 . . .9 '.'. 13'.2 '.'. 14*.0 ..1.0 .. .9 14.4 .8 15.4 1.6 15.6 1.0 17.2 1.0 •• 16.2 16.9 .7 .5 17.3 18.2 .9 1.2 White Av A.D Col. Av A.D Newport News. .. .. 12.8 12.6 .. ..1.0 .8 .. 13.0 .. 14.3 ..1.0 .. .8 13.6 .6 •• •• •• .. .. 36 THE PSYCHOLOGY OF THE NEGRO. TABLE IV. Ages — Difference in Years Between the White and the Colored Subjects Tested (Minus signs indicate that the colored subjects are of greater age.) Richmond. Grs.5A 5B 6A 6B 7A 7B lA 2A 3A 4A Av. P.E. .1 _,3 .7 —.3 —.7 —.9 _ .5 — .8 — .3 — .6 — .36 .09 Fredericksburg. , . —.7 . . —.4 . . —1.0 —1.6 —1.1 —1.3 —1.02 .11 Newport News. .... .. —1.7 —1.7 Mayo, Phillips and Pyle, (see Chapter I), also found that the colored children tested by them were appreciably older than white children of the same school grade, and this is indicated by school censuses in general, (see Mayo, '13). This age difference is significant as showing that colored pupils are less advanced than white pupils in school work. But it is not important in a comparison of the standing of the two races in mental tests, if the comparison is made by both age and grade. When the scores are compared by ages, the white pupils in the comparison will be of higher school grade; when they are compared by grades, the colored pupils will be of greater age. In the former case any possible ad- vantage will be on the side of the whites; in the latter case it will be on the side of the colored pupils. If success in the tests is not dependent upon school training, it is obvious that an age comparison is the better. But if success in the tests does depend upon school training, a comparison by grades is to be preferred. Where the influence of school training upon standing in the tests is unknown, a comparison by both age and grade would seem to be advisable. In this mono- graph the results in all of the tests are set forth by both ages and grades, but evidence will be brought forth to show that ability to perform the tests is not appreciably dependent upon school training, and consequently the age comparisons are the more reliable. A comparison by ages, indeed, is to be pre- ferred to one by grades in all serious investigations of racial mental differences. For a grade is essentially a group of people selected because they are much alike in capacity. And a mental test which is dependent upon academic training THE SUBJECTS AND THE TESTS. 37 must be comparatively worthless as an index of true racial ability. The Tests The tests employed in this investigation were selected pri- marily with a view to ascertaining racial differences in the higher rather than in the lower intellectual capacities. It is in the higher capacities that men are supposed to differ most. And it is these capacities that are of greatest influence in determining their relative achievement. The investiga- tions previously made and the views previously held indicate that there are no considerable group differences in sensation, in motor control, in native retentiveness. The differences to which evidence has pointed have been, on the side of intellect as opposed to feeling, in such abilities as those included under the terms constructive imagination, the apprehension of mean- ing, reasoning power. These latter traits divide mankind into the able and the mediocre, the brilliant and the dull, and they determine the progress of civilization more directly than do the simple and fundamental powers which man has in com- mon with the lower animals. While testing these traits, it was thought advisable at the same time to employ certain tests of lower capacity for the sake of comparison. The tests used were the Woodworth and Wells Mixed Relations, I and II; a form of the Ebbinghaus Completion Test; a Cancellation Test; and one of the Columbia Maze Tests. A test of immediate memory was also given, but the results from it were discarded. In this test the series of digits to be remembered were presented orally by the ex- perimenter, and some of the pupils, both white and colored, undoubtedly wrote down the numbers surreptitiously as they were called, instead of waiting until after the series were finished. It is interesting to note that evidence of this occurs more frequently in the results from the lower than from the higher grades, and more frequently in the colored than in the white schools. A possible "study of dishonesty" is suggested. The mixed relations test, in its various forms, has been used by a number of investigators, and has been highly recom- mended as a test of intelligence. Wyatt ('14), in a study undertaken in order to determine reliable intelligence tests, found correlations of .80 and .62 between the mixed relations test and careful subjective estimates of intelligence, in two 38 THE PSYCHOLOGY OF THE NEGRO. groups of subjects. He also found a correlation of .85 be- tween the mixed relations and completion tests, and states that these two tests correlate more highly with inteUigence than do any of the thirteen other tests employed by him, and that they also correlate more highly with the other tests as a whole. Vickers and Wyatt ('13) attempted to deter- mine suitable tests for assigning children to school grades, and found the mixed relations, completion and hard opposites to be most satisfactory. The correlations between the mixed relations test and intelligence, (inteUigence being defined as adaptability to new conditions), were .51, .61, .64 and .86 with four different classes of children. The reliability of the test, as m.easured by its correlation in repeated trials, was high, the average correlation between the trials being .70. Burt ('11) found a correlation of .52 with intelligence, and a coefficient of rehability of .92. He recommends the test, along with the completion test, as being an excellent indica- tion of ability in logical inference or reasoning. The particular form of the test as used is the one de- signed and standardized by Woodworth and Wells ('11), its originators. It consists of two parts, numbered I and H, and is in reality two separate tests. The parts are of equal dif- ficulty and the twenty relations in each part differ as little in difficulty as it was possible to make them. The test is printed in full in the appendix, as are the others used in this investigation. It is always difficult to state just what mental function is experimented upon by a given test. The various traits so overlap and are so dependent upon one another in their action that no one trait can be completely isolated. If a test cor- relates well with other tests of the same or related perform- ances, it may be taken as a reliable index of ability in the functions involved, although the specific functions themselves cannot be definitely and exhaustively described. Most tests are so regarded. But within limits it is possible to state approximately the functions that are tested. The mixed relations test is primarily one of controlled association of the sort that is the basis of all efficient reasoning. It demands that a relation be perceived and applied; and then that an- other and a different relation be perceived and applied, and so on through the test. A simple test of controlled association, such as the opposites test, requires that a mental set or THE SUBJECTS AND THE TESTS. 39 determining tendency be previously formed at the instance of the experimenter, and then apphed unchanged to stimuli as they occur. The mixed relations test goes further than this, and requires that the mental set itself be determined by the subject before it is applied, and that a different mental set be fixed for each stimulus. It involves what James ('92) has called "sagacity", and requires association by similarity, the perception of meaning, voluntary control of ideas. Wood- worth and Wells state that the test measures skill in hand- ling associates by means of a determining tendency; mental alertness and flexibility; that it is a "logical relations" test. It is safe to say that the functions involved are those most intimately concerned in that successful handling of material which distinguishes the intelligent and mentally active indi- vidual from the unintelligent and dull. The language factor is of course present, but where the words used are simple and well known, this is not important in a test of individual or group differences. (See Simpson, '12, p. 69ff.). The completion test was invented by Ebbinghaus ('97), and has been widely used. Wyatt ('14), as mentioned above, recommends the test as one of intelligence; he finds correla- tions of .85 and .61 with subjectively estimated intelligence in two classes of school pupils. Vickers and Wyatt ('13) found correlations of .82, .88, .76 and .82 with intelligence in four groups of subjects. And they found correlations of re- liability of .84, .87, .66 and .69 in successive trials with the four groups. Burt ('11), using two forms of the test, found correlations with intelligence of .48 and .53. Brown ('11) states that the test correlated .43 and .69 with general intel- ligence in two groups of subjects. Simpson ('12) found a correlation of .92 for reliability. Correlations between the completion and a number of other tests used by him were: hard opposites, .92; easy opposites, .75; memory of words, .92; memory of passages, .91; cancellation, .68; adding, .71; geometrical forms, .54; learning pairs, .72; completing words, .50; drawing lengths, .26; estimating lengths, .52. A cor- relation of .67 between the completion test and the average of nine other varieites of association tests was reported by Whitley ('11). The form of completion test herein used is composed of the twenty-five separate sentences which constitute sentences 23 to 47, inclusive, of a completion test designed by Mr. M. R. 40 THE PSYCHOLOGY OF THE NEGRO. Trabue*, of Columbia University. These sentences are grad- uated in difficulty, and have been standardized by Trabue in experiments upon several thousand school children. Numbers 23 to 47 were selected because they were well adapted in diffi- culty for use with the school grades to be tested. A comple- tion test in the form of separate sentences has several advan- tages over one in paragraph form, such as has generally been used. Where a paragraph is employed, accidental factors are much more likely to influence the result. The subject of the paragraph may be relatively unfamiliar to any given pupil; or one unusually difficult part of the whole may render the completion of a large section of the rest unduly difficult. This cannot occur where there are a number of separate sentences dealing with different subjects and each counting as a unit. Then, too, in scoring it is probably better to score on the basis of sentences than of words. Thought proceeds by judgments, whole sentences, not by words. And in a test of thought power rather than of language power one should be able to gauge the apprehension of meaning as a whole rather than what is perhaps the more distinctively literary ability re- quired to fit a word into a specific context. It is thought, not its vehicle, that it is to be measured. The ability to use language must probably remain a factor in this test, though not the most important factor where the material is familiar. Thought and language are largely implications of each other, and in great measure ability in one means ability in the other. But as far as possible the language factor should be elminated and the thought factor emphasized. The mental functions measured by the completion test are akin to those involved in the mixed relations test. "This is in- dicated by the high correlation between the two. Ebbing- haus ('97) described it as essentially a test of intelligence, requiring the ability to combine separate impressions into a coherent whole. Simpson ('12) calls it a test of selective thinking. Whitley ('11) classes it among her association experiments. It would seem that the ability to perceive rela- tions, to apprehend meaning, to control association in order to fill a gap, is implied in a successful performance of the test. Association or selective thinking or intelligence are perhaps equally good terms to apply to the processes. In the language *Since the above was written a preliminary account, (Trabue, '15), of this test has been published. THE SUBJECTS AND THE TESTS. 41 of popular speech, the test requires "good sense" and a "quick mind". The maze test employed is the so-called straight maze de- signed and used at Columbia University. This has been found (see Whitley, '11) to be the most satisfactory form of maze test. It is comparatively easy to score, and the eye strain resulting from its use is negligible as compared with certain other forms of maze. A correlation of .49 was found by Whitley between the straight and the average of three other varieties of the test. Simpson ('12), using a scroll maze, found a coefficient of reliability of .76 for the test and a co- efficient of correlation of .26 with the average of twelve tests of intellectual functions. The traits measured are quickness and accuracy of move- ment in drawing a line between the two sides of the maze without touching them — motor as opposed to intellectual abilities. The concellation test is the familiar "A Test", designed by Cattell and Farrand ('96). It is a regulated pied text, and contains one hundred A's and sixteen of each of the other letters — five hundred capital letters in all. The test has been used by many investigators and in a variety of forms, such as the "A— T" test, the "E— R" test, etc.; and it has been described as a test of various functions. Pillsbury ('08) says it is one of the best tests of degree of attention. Whipple ('10) agrees that this is one of its main features. Bourdon ('95) used it to measure discriminative ability. Judd ('07) similarly classes it as a test of discriminative reaction. Cattell and Farrand ('96) regarded it as a test of rate of perception. Thorndike ('04) also used it as a test of perception. Pyle ('13) gives it as a test of perception and attention. On the whole, it seems that the test measures all of these capacities as they function together. To cancel A's it is necessary to perceive them discrimnatively and attentively, and to react by the simple cancelling movement. The correlations of the test with other tests and with class standing are generally small and sometimes negative. (See Whipple, '10). In giving the tests, the instructions to the subjects were by means of examples on a blackboard, supplemented by such oral directions as were necessary. A constant order of suc- cession was maintained among the tests. The maze was given 42 THE PSYCHOLOGY OF THE NEGRO. first, while the pencils were sharp, then followed cancellation, mixed relations I, mixed relations II, and completion. This order was invariable except in Newport News, where, on ac- count of a lack of time, the mixed relations and completion tests were given first, to make sure that they, as the most important, would be finished. The schools were taken as wholes, one after another, and the lower grades were generally tested before the higher. There was no appreciable chance for aid in working the tests to be transmitted in conversa- tion from grade to grade. The time-limit method was used, and the effort was made to allow just enough time in each test to enable to quickest of all the subjects to finish. In Fredericksburg, where the tests were given first, the time allowed for the mixed relations test was 15 sec. longer than elsewhere, and the time allowed for the completion test was 30 sec. longer. The time was reduced in the other cities because several of the Fredericks- burg subjects finished the work before the time limit was reached. This extra time in Fredericksburg had apparently no effect upon the relative standing of the white and colored groups, but it is obvious that if the time allowed is longer than is necessary for the quickest subjects to complete the test, the group differences will be somewhat reduced. The brightest of the pupils will not be able to accomplish as large an amount of work as their ability warrants. On the v/hole, the time limits as used were approximately the periods required by the ablest of the subjects. With the exception mentioned, these times were as follows, for all grades: each mixed rela- tions test, 1 min., 45 sec. ; completion test, 8 min., 30 sec. ; maze test, 1 min., 30 sec. ; cancellation test, 1 min., 20 sec. The sub- jects were told that they would have barely time to finish if they worked at their highest speed. In all of the tests the directions were to try for as great speed as possible, while not making any mistakes. If an unusually difficult part of the test should be met, in the mixed relations and completion tests, they were advised to pass it by without a too great waste of time. A stop-watch was used, and all directions were given by the writer, who also did all of the scoring. In scoring the mixed relations test, each accurate relation recorded was graded 2; each partially correct relation was graded 1 ; each error was graded 0, as was each omission. The THE SUBJECTS AND THE TESTS. 43 possible maximum score was thus 40 for each of the two parts in which the test was given. The scoring was, of course, absolutely uniform throughout, for each type of correct, partially correct or incorrect record. The completion test was scored in the same way. Each correct sentence was rated 2; each partially correct sentence was rated 1 ; and each incorrect or omitted sentence was rated 0. Since there were twenty-five sentences, the possible maxi- mum score was 50. The maze test measured two things, speed and accuracy. Accuracy was scored by counting the number of touches made. Speed was determined by the amount of the test completed or the distance traversed. In rating this, each straight sec- tion of the maze was counted 1. Since there are 140 straight sections in the test, the possible maximum score for speed was 140. This test presents a difficulty in its scoring. The two quantities which it measures are variables which do not main- tain a constant ratio to each other. If only a short distance is traversed within the time limit, the number of touches is small as compared with the amount done. But if a great distance is traversed, the number of touches is large as com- pared with the space gone over. To illustrate: a distance of 60 will mean, say, 4 touches, a ratio of 1 to 15 ; but a distance of 120 will mean, say, 20 touches, a ratio of 1 to 6. And yet the latter record may be as good as the former. As speed increases, accuracy normally decreases, and in a constantly changing ratio. In handling the maze test, a number of investigators have chosen to deduct a certain arbitrary amount from the speed record for each touch. Thus Whitley ('11) and Simpson (*12), using the am.ount-limit method, add 5 sec. and 10 sec, respectively, for each touch made. But this cannot be satis- factory. A touch made by a subject who works at great speed is far less significant than a touch made by a subject who works at a much lower speed. Where the ratio of accuracy to speed is a variable one, as in this case, no constant figure can be deducted from the speed record for each error. What is needed is a set of ratios, expressing the relation of touches to distance at each of a large number of possible distances. Such a set of ratios could be ascertained by ade- quate experiment, and they would be very interesting in 44 THE PSYCHOLOGY OF THE NEGRO. themselves and also very useful, since there is no other prac- ticable group test of motor capacity. But at present they do not exist, and it is therefore impossible to give the best treat- ment to the results of the maze test. The method that is followed herein is to set forth the num- ber of touches and the distance covered by each group of sub- jects. The comparison of group with group must then be based on the relative number of touches made and the relative distance traversed. If the score for either touches or distance, or both, should be the same for each group, an accurate com- parison would of course be possible. Or if one group should exceed the other in touches but not in distance, or in distance but not in touches, an accurate comparison could be made. It generally happens, however, that one group exceeds the other in both touches and distance, and in this case the difficulty arises. The ratio of touches to distance is normally smaller in a slow group than in a fast one, and smaller to an unknown degree. The cancellation test also measures both speed and ac- curacy, but here the difficulty in equating the two is not so pronounced. The score for speed is obtained by counting the number of A's cancelled ; the score for accuracy is arrived at by counting the number of A's omitted in the amount of text gone over. The number of A's omitted is very small, and is constantly so from group to group. Roughly, only about one-third or one- half of the subjects make any omissions at all. And it does not appear that there is any very definite relation between the number of omissions and the number of cancellations. Sev- eral investigators have reported this to be true of the test. Woodworth and Wells ('11) say there are no workable indi- vidual differences in accuracy and that there is little reason for its being scored. Binet ('03) found that subjects worked with approximately equal accuracy. Others, as Thorndike ('04), ignore omissions in giving results from the test. Whipple ('10) and others, however, think omissions may be important. Both omissions and cancellations are given for all of the groups in this study ; but it will be apparent that signifi- cance can be attached only to the latter. Of the tests used, the mixed relations and the completion were given to all subjects. The maze and the cancellation THE SUBJECTS AND THE TESTS. 45 were not employed with the white high school pupils of Rich- mond, on account of a lack of available time. The maze was also not used with the white elementary pupils of Newport News. The white elementary pupils of Fredericksburg, through an oversight on the part of the experimenter, did not sign their names to the maze test, and their records in this test can consequently be used only in grade and not in age and sex comparisons. The total time consumed by these tests in the case of any individual subject was small, and it cannot be supposed that the records obtained from a given individual are an accurate index of his relative ability in the traits measured. Further trials would be necessary to establish the final standing of any one of the pupils in the tests. But while this is true, it is also true that the tests are sufficient to establish with accuracy the relative standing of large groups of subjects as wholes, and it is a group — a racial — comparison that is in question. Where a group is tested, the chance inaccuracies which deflect the true position of one individual in one direc- tion serve also to change the position of another individual in the opposite direction, so that on the whole the central ten- dency of the group remains unchanged from what it would be in the case of a very great number of measurements, the practice effect, of course, being disregarded. The individual inaccuracies balance each other, and the group standing is un- affected by them. It is improbable that many repeated trials would appreciably disturb the average score. This view is taken for granted in all tests upon large numbers of subjects ; Thorndike ('04) gives illustrations of its validity. The mixed relations test as herein used is another illustration, as will appear. For the test is given in two parts — is, in reality, two tests — and the relative standing of the groups compared is the same in each. Woodworth and Wells ('11), indeed, recom- mend a short test of this and related kinds as being better than a long one, since it is freer from interferences of a disturb- ing character. So the final average results of these tests may confidently be taken as reliable measures of the relative ability of whites and negroes in the traits involved, although the time consumed in the actual testing was short. CHAPTER III GENERAL COMPARISON OF WHITES AND NEGROES In the present chapter a general comparison is made be- tween the scores of the white and the colored subjects. In set- ting forth the results, the tables and graphs are arranged in the same order for each test. First are given the average scores made by each age and sex and by each grade, with their aver- age deviations. The tables containing these data are the basis of the comparisons made in the other tables, and they also give the figures which are represented by the graphs. The graphs are next in order, and serve to make the group relationships contained in the tables somewhat plainer. It may be noticed that the graphs and the tables of comparison which follow them omit those ages and grades in which there are not fig- ures for both races : no comparison is made unless the ages and grades are alike. The results from Richmond, Fredericks- burg and Newport News appear in succession in each table. There are no graphs for the two latter cities. Following the graphs, for each test come the tables in which the group comparisons are made. First appear the actual differences between the scores obtained by the two races, classified by age and sex and by grades. The averages of the separate age and grade differences are given, with their probable errors. These averages, of course, are the most reli- able figures for comparison. Next are the tables which set forth the percentage of the score of the whites which was obtained by the negroes, with the averages of the separate ages and grades and the probable errors of the averages. The tables which show the actual differences between the scores and those which show the percentages deal with the same racial differences, and are simply two ways of exhibiting the same group relationships. Lastly, as still another mode of comparison, for each test is given the percentage of each age and grade of the colored subjects that reaches or exceeds the 46 COMPARISON OF WHITES AND NEGROES. 47 average score of the white subjects of the same age or grade. The averages and probable errors of the separate age and grade percentages are given as before. This comparison is made only for Richmond, since the relatively small number of pupils tested in Fredericksburg and Newport News would ren- der it somewhat unreliable for those cities. A group com- parison by means of the percentage of one group reaching or exceeding the average of the other requires a rather large number of subjects to be of value. If the number is small, chance inequalities in the distribution of the groups may make the percentage appear to be much too high or much too low. It should be said that this comparison in terms of the per- centage of the negroes reaching or exceeding the average of the whites is valid only in so far as the groups follow a normal mode of distribution. That is, the average of the groups must also be approximately the median of the groups. Or to put it in another way, approximately fifty per cent, of the groups must reach or exceed their own average. Otherwise, we would have fifty per cent, of the negroes reaching or exceeding the score obtained by 30 per cent, or 70 per cent., or any indefinite per cent., of the whites. The ideal procedure would be to ascertain the percentage of the negroes reaching or ex- ceeding the median, rather than the average, of the whites. But in the present comparison the averages and the medians of the groups compared are approximately the same, as the following figures will show, and the mode of comparison adopted is a valid one. In the mixed relations test, the percentage of the whites reaching or exceeding their own average in Test I is 50.0, P.E. 1.7, for all ages of the boys ; 53.5, P.E. 2.2, for all ages of the girls; and 54.8, P.E. 1.6, for all grades. (The grades contain both boys and girls). In Test II the percentage for boys, all ages, is 55.0, P.E. 2.3; for girls, all ages, it is 55.9, P.E. 2.2; and for all grades it is 55.6, P.E. 2.1. It is thus evident that for the white subjects in this test the average and the median are approximately the same, and that the groups follow a sufficiently normal form of distribution for the percentage comparison to be made in terms of the average rather than of the median. The figures showing the percen- tage of the colored subjects reaching or exceeding their own average are as follows : Test I — boys of all ages, 48.4, P.E. 2.4 ; 48 THE PSYCHOLOGY OF THE NEGRO. girls of all ages, 48.7, P.E. 2.3; all grades, 48.5, P.E. 1.4. Test II_boys of all ages, 52.1, P.E. 2.5 ; girls of all ages, 51.2, P.E. 3.3 ; all grades, 52.0, P.E. 1.9. Thus the colored subjects are also seen to be so distributed that their central tendency is approximately the same whether measured by average or by median. It worthy of note that the percentage of subjects reaching or exceeding their own average is somewhat higher in Test II than in Test I. This is due to the fact that the scores themselves in Test II are higher than in Test I, on account of the practice effect of Test I, to such extent that several of the brighter subjects finished the second test be- fore the time limit was reached, and thus did not attain their possible maximum score. This lowers the average standing in Test II so that more than fifty per cent, of the subjects reach or exceed it. This occurs for both races, though slightly more so for the whites than for the negroes, since the score of the whites was greater, as will appear, and more of them finished the test before the expiration of the time limit. A slightly larger percentage of the negroes, as will also appear, reached the average of the whites in Test II than in Test I. But this should not be taken to indicate that the negroes profited more by the practice in the first test than did the whites: both races reached or exceeded their average in greater numbers in the second test. In the completion test, the percentages of white subjects reaching or exceeding their own average are: boys, all ages, 50.2, P.E. 2.0; girls, all ages, 52.2, P.E. 1.3; all grades, 52.7, P.E. 1.2. The percentages of colored subjects reaching their own average are: boys, all ages, 52.6, P.E. 2.6; girls, all ages, 54.2, P.E. 2.1 ; all grades, 54.4, P.E. 1.3. In this test, as in the mixed relations, the average and the median are approxi- mately the same. The maze test shows the following percentages of white subjects reaching or exceeding their own average: boys, all ages— Touches, 50.0, P.E. 3.3, Distance, 52.0, P.E. 1.6; girls, all ages— Touches, 52.4, P.E. 2.4, Distance, 46.8, P.E. 2.8; all grads— Touches, 48.2, P.E. 2.7, Distance, 49.0, P.E. 2.9. The percentages of colored subjects reaching or exceeding their own average in the maze test are: boys, all ages — Touches, 46.6, P.E. 1.9, Distance, 46.4, P.E. .9 ; girls, all ages- Touches, 45.6, P.E., .9, Distance, 44.4, P.E. 1.8; all grades- Touches, 47.8, P.E. 2.1, Distance, 51.2, P.E. 2.3. COMPARISON OF WHITES AND NEGROES. 49 In the cancellation test, the percentage of a group reaching its own average or the average of another group can be validly computed only for the cancellations, not for the omissions. Less than half of the subjects made any omissions at all, and the distribution of the omissions is consequently very assymmetrical. The percentages of whites reaching or exceed- ing their own average in cancellations are as follows: boys of all ages, 53.0, RE. 1.3; girls of all ages, 49.6, P.E. 1.9; all grades, 49.3, P.E. 1.4. The percentages of negroes reaching or exceeding their own average in cancellations are: boys of all ages, 50.0, P.E. 3.6; girls of all ages, 52.2; P.E. 3.7; all grades, 53.7, P.E. 1.5. These figures make it evident that we can compare the white and colored subjects in terms of the percentage of the colored reaching or exceeding the average of the white, since the distributions throughout are on the whole symmetrical.* Perhaps something should also be said as to the other tables of comparison, those which show the actual differences between the scores of the two races and those which show the percentage of the score of the whites that is obtained by the negroes. It sometimes happens that investigators average the scores of a number of different ages or grades, and make comparisons by exhibiting the relation between such averages. But this procedure makes it impossible to compute the proba- ble error of the differences, and thus renders the comparisons doubtful. If, for example, scores from ages 11 to 18 are averaged for two races, and a certain difference between the averages is set forth, one cannot tell from this difference alone whether any real racial superiority has been found. *Note may be made of the fact that in a few ages and grades no colored subjects reached or exceeded the average of the whites, and that the percentage in such cases was put down as zero. But such per- centages are really less than zero. And when they are counted as zero, as they are, in obtaining the average for all ages or grades, it is obvious that this procedure tends to make the average percentage too high. That is, it tends to make the racial difference appear to be less than it really is. Another matter that also makes the racial differences found appear to be less than they really are, is the fact that in the tables which show the percentage of the score of the whites obtained by the negroes, the per- centages are always expressed in terms of the white score. Hence whenever the negroes have a higher score than the whites, as they do in a few instances, the percentage expressing the fact is dispropor- tionately large. For example, if the whites score 100 and the negroes score 75, the percentage is 75; but if the whites score 75 and the negroes score 100, the percentage is 133. These considerations affect the final average-s only slightly, but they should be borne in mind. 50 THE PSYCHOLOGY OF THE NEGRO. It may be that the separate ages show a superiority first in favor of one race and then in favor of the other, and so on. What is needed to determine the validity of the difference is its probable error. This will indicate how constant the dif- ference is from age to age — whether it is a true difference or is due to chance. Of course it will be recognized that probable errors can be computed for the averages of a number of dif- ferent age or grade scores. But such probable errors will be larger than the racial variability from age to age would war- rant, since scores normally increase with age. And they therefore cannot be used as accurate measures of the probable error of the racial difference. Where, as in the tables of comparison contained herein, the differences between the scores of the separate ages are found, the average of these differences can be obtained and also the probable error of the average. If the differences for the separate ages are pronouncedly in favor of one of the two races, the probable error will be small as compared with the average. Even though slight, the average difference will in- dicate a real racial superiority if its probable error is small. But if the differences between the separate ages are chance differences which favor first one race and then the other, no matter if the average difference is large, it will yet be seen to be no true index on account of its large probable error. This principle is as applicable to those of the following tables which exhibit the percentage of the score of the whites ob- tained by the negroes, as it is to those which set forth the actual differences. In the percentage tables also, the average of the different ages is computed and its probable error is given with it. Mixed Relations Test The scores obtained by the two races in this test are shown in Tables 5 and 6, and in Figures 1-6. It may be noted that in Test II the scores are somewhat higher than in Test I. This is evidently due to the practice that the first test afforded. It may also be noted, (Figs. 3 and 6), that there is a pro- nounced jump in the scores of the colored grades when high school is reached, and that this jump does not occur in the scores of the white grades. This is probably the effect of the more intense action of selection upon the colored than upon the white high school pupils, as was pointed out in the preceding COMPARISON OF WHITES AND NEGROES. 61 chapter. In some of the high school years the colored are superior to the white subjects, but this does not occur in the elementary school. TABLE V. Mixed Relations Test — Scores by Age and Sex Richmond. Ages 10 11 12 13 14 15 16 17 18 19 Test I. Boys — white Av 14.8 16.5 18.3 18.0 20.1 27.8 25.6 25.0 A.D 6.6 7.3 8.7 7.3 10.0 5.9 8.1 4.0 Boys— Col. Av 11.5 10.2 11.8 13.7 12.9 20.5 22.0 20.0 15.3 A.D 3.8 4.4 5.0 6i5 6.9 8.4 8.5 8.4 7.6 Girls — white Av 16.9 18.8 18.0 22.7 23.0 25.0 19.8 16.6 18.5 A.D 6.0 4.2 7.4 8.3 8.6 7.4 6.2 3.8 .5 Girls— Col. Av 10.0 12.4 10.6 12.4 14.5 18.6 21.1 26.0 18.4 22.6 A.D 5.3 4.4 4.8 6.0 7.5 9.1 6.8 6.8 8.4 9.3 Test II. Boys — white Av 16.5 19.5 21.3 24.1 22.0 33.0 34.6 29.6 A.D. 8.0 9.3 9.2 9.5 11.9 5,5 5.2 4.6 Boys— Col. Av 12.6 11,6 13.7 16.3 19.2 25.7 32.0 32.1 31.2 A.D 8.1 6.6 5.8 10.0 11.0 8.5 7.1 7.1 3.7 Girls — white Av 22.0 22.4 21.5 24.7 28.6 31.7 27.« 23.0 18.5 A.D 6.4 8.2 8.5 10.7 7.4 6.9 7.2 11.4 1.5 Girls— Col. Av 14.0 17.2 10.6 11.8 16.4 22.1 27.9 31.9 22.7 24.0 A.D 4.0 5.0 4.7 6.6 10.8 9.1 6.8 7.1 9.2 5.3 Fredericksburg. Test I. Boys — whit© Av 10.6 20.5 14.4 23.1 27.5 23.5 28.8 38.0 A.D 4.0 12.5 6.4 10.7 9.4 13.1 8.5 2.0 Boys— Col. Av 11.3 9.5 .. 14.0 11.2 .. 22.0 19.5 A.D 1.6 3.5 .. 5.3 2.6 .. 8.5 Girls — white Av 20.5 19.5 22.5 19.3 16.6 26.8 20.5 A.D 11.1 6.1 11.1 14.0 8.0 9.6 10.5 Girls— Col. Av 10.7 10.5 6.0 28.5 21.3 13.5 18.5 A.D 4.7 1.9 4.0 8.5 8.9 6.5 7.5 Test II. Boys — white Av 22.6 23.5 17.8 26.5 33.4 32.1 34.0 38.0 A.D 6.6 15.2 10.8 12.5 7.5 6.5 6.8 1.3 Boys— Col. Av 13.3 8.5 .. 20.3 19.0 .. 29.0 22.7 A.D 5.0 .5 .. 9.6 7.6 .. 2.0 13.2 Girls — white Av 25.8 19.1 29.0 22.6 24.8 32,2 28.0 A.D 12.1 7.1 9.7 8.3 5.5 7.0 8.0 Girls— Col. Av 13.3 11.0 16.0 35.0 27.6 23.0 21.0 A.D 8.1 2.8 12.0 3.0 5.8 9.0 16.0 52 THE PSYCHOLOGY OF THE NEGRO. Ages Test I. Boys — white Av A.D Boys— Col. Av A.D Girls — ^white Av A.D Girls— Col. Av A.D Test II. Boys — white Av A.D Boys— Col. Av A.D Girls — white Av A.D Girls— Col. Av A.D 10 Newport News. 11 12 13 14 15 16 27.6 16.4 20.9 18.9 15.2 3.3 6.4 13.5 9.5 8.3 5.3 7.0 1.0 4.7 •• 19.6 23.1 18.7 16.3 10.7 24.5 4.6 6.6 7.3 7.1 4.7 1.5 7.7 7.0 6.0 2.0 8.6 •• 4.1 1.0 1.8 6.6 83.6 24.4 24.7 20.8 21.0 3.3 7.1 21.5 15.5 9.4 8.3 7.0 1.0 5.2 17 18 19 30.4 30.9 23.9 22.0 12.5 31.0 2.4 5.6 7.8 10.2 10.0 3.0 9.4 3.0 4.2 2.6 6.8 2.4 4.6 3.3 4.0 1.3 TABLE VI. Mixes) Relations Test — Scores by Grades Richmond. Grades 5A 5B 6A 6B 7A 7B lA 2A 2B 3A 4A Test I. White Av 13.1 13.9 16.8 16.8 18.4 4.0 4.6 6.2 5.8 24.9 6.9 25.5 7.3 28.0 6.8 •• 23.1 23.3 A.D ... 6.5 7.2 8.6 Col. Av . . . . 8.9 10.3 13.0 13.4 9.7 12.2 21.4 20.7 17.7 29.4 22.6 A.D ... 3.5 3.1 5.9 5.2 5.1 6.3 6.8 7.7 6.1 4.4 9.3 Test II. White Av . ... 14.8 18.0 19.8 22.6 21.8 29.4 29.0 33.6 31.8 30.2 A.D ... 7.0 6.2 9.2 7.5 9.1 8.5 7.9 5.2 6.9 7.8 Col. Av . . . . 9.7 10.3 14.0 13.9 16.5 11.8 27.1 28.1 24.6 35.4 29.6 A.D .... 4.8 5.6 6.7 7.1 8.2 6.6 8.5 6.9 8.0 2.6 8.2 Test I. Fredericksburg. White Av..... • • • • • • .. 16.2 .. 18.5 23.1 22.6 26.3 30.8 A.D • • • • • ..8.1 .. 10.0 9.8 12.8 9.8 7.1 Col. Av > • • • • • .. 10.7 .. 8.2 13.6 16.2 22.5 24.2 A.D ... ..2.7 .. 4.5 4.8 6.2 9.8 6.2 Test II. White Av • • • • • • .. 19.1 .. 22.0 32.5 27.8 32.9 34.3 A.D • • . .. 10.1 .. 11.0 6.7 8.8 5.2 6.9 Col. Av • • • • > • .. 10,6 .. 11.1 20.6 23.5 30.3 30.1 A.D .. 3.8 .. 4.7 9.0 6.1 7.8 7.1 Test I. Newport News. White Av • • • • • • . . . . 15.7 22.5 20.6 . . A.D • • • • • .. ..6.8 6.1 7.3 . . Col. Av • * • • • ..7.8 .. 7.4 . . A.D • • • • • ..3.5 .. 3.6 . . Test II. White Av . . . . 23.2 24.9 25.5 A.D • • * • • .. ..9.1 8.2 8.6 .' .' Col. Av • • • • • ..8.5 .. 7.1 . . A.D .. 4.3 .. 3.8 , , , , . . . • COMPARISON OF WHITES AND NEGROES. 63 SCORE JO /J /o 6 -1 ^ u i] 1 I n 1 /\&E W / c 1 w c / 1 /2 W C 1 w c |w cl w c Iw c 1 /6 \ /7 \ w cl /r 1 Fig. 1. Mixed Relations Test I — Scores of White and Colored Boys — Richmond.* *Th€ white and the black columns indicate the scores of the white and the colored subjects, respectively. scone SO 2S (20 /s 10 5 W V 1 1 1 ( \ t 1 AGE w c 1 w // 1 / c Iw 2 1 / w 71 w / c 1 w s 1 / 71 w c Iw / 7 1 / 71 Mixed Relations Test I — Scores of White and Colored Girls — Fig. 3. Mixed Relations Grades — Richmond. Test I — Scores of White and Colored 54 THE PSYCHOLOGY OF THE NEGRO. Fig. 4. Mixed Relations Test II — Scores of White and Colored Boys- Richmond. rvv c // Fig. 5. Mixed Relations Test II — Scores of White and Colored Girls- Richmond. Fig. 6. Mixed Relations Test II — Scores of White and Colored Grades — Richmond. COMPARISON OF WHITES AND NEGROE^S. 55 It is interesting to note that the gradual increas-.e in the scores as the age of the subjects increases comes to a stand- still in the upper years; indeed, the graphs show thad the older pupils as a whole have somewhat lower scores than tK^ose immediately below them in age. This is to be expected. In a group of pupils taken from any limited number of school grades, those whose ages are highest will not do as well in tests of ability as their ages would seem to warrant. When the elementary school pupils alone are classified by age their scores do not progressively increase, but instead tend to de- crease in the upper years as do those of the high school pupils represented in the graphs. The white elementary pupils tested in Richmond ranged, in numbers large enough for compari- son, from 11 to 15 years of age, inclusive. And their scores for each age were as follows: Mixed Relations Test I — Boys, 14.8, 15.0, 17.3, 16.0, 13.7; Girls, 16.9, 18.8, 17.0, 17.2, 15.2. Mixed Relations Test II— Boys, 16.5, 18.0, 20.6, 22.1, 15.0; Girls, 22.0, 22.1, 21.0, 17.4, 16.7. The scores for the dif- ferent ages of colored elementary pupils behaved in the same way. It is thus evident that there is precisely the same sort of decrease in the scores of the higher ages that was found for the high school pupils. The explanation of this decrease with age is probably to be sought in the bearings of the general fact that within any school grade the younger pupils have greater natural ability than the older. If the older pupils had had great ability, they would have passed out of the grade; that the younger pupils are so advanced as to be classed with the older pupils is evidence of their considerable capacity. The younger pupils in a shorter length of time have done the same amount of work that the older pupils have done in a greater length of time. A grade contains the best of the young and the worst of the old. So when the pupils in the upper grades of the elementary school are classified by age, it is apparent that the older pupils will be those of less native ability and the younger pupils will be those of greater native ability. The younger pupils in the upper elementary grades are the best of their age; the older pupils are the poorest of their age. The poor and mediocre pupils of the same age as the younger group have not yet reached the higher grades of the elementary school; the able pupils of the same age as the older group have passed on into high school. The same V 56 / THE PSYCHOLOGY OF THE NEGRO. reasoning- applies to the school system as a whole, including the hi'^h school as the highest grades. The older pupils in the upper grades of the high school have not as great natural ability as the younger pupils. The ablest pupils of the same age as the older group have finished the high school course, and left behind only the mediocre and the poor. But the younger group is composed of the brightest pupils of their age, for those of less ability have not yet reached the upper grades. That this explanation of the smaller scores made by the oldest pupils tested is pertinent, is further indicated by the fact that the graphs which give the scores by grades do not show the decrease for the advanced pupils as it is shown by the graphs which give the scores by ages. Grades are supposedly groups of an increasiing degree of ability. The truth of the view set forth above is still further borne out by an incidental comparison that came to light in the present investigation. In the elementary schools at Richmond were tested 30 white boys, 20 white girls, 24 colored boys and 38 colored girls, 112 in all, of ages 14 and 15, combined. In the high schools at Richmond were tested 18 white boys, 25 white girls, 10 colored boys and 31 colored girls, 84 in all, of ages 14 and 15, combined. It is true that there was a larger proportion of 14 than of 15 year old pupils in the elemen- tary groups, and a larger proportion of 15 than of 14 year old pupils in the high school groups. But this is of no conse- quence, since the 14 year old elementary pupils are fully as able as those 15 years old, as was shown two paragraphs above; and since the same thing is true of 14 and 15 year old pupils in the high school, as will be shown three paragraphs below. So these groups may be considered as of the same age, one set being in the elementary school and one set in the high school. The scores obtained by them in three tests were as follows: Mixed Relations Test I — White Boys, Elementary, 14.8, High School, 24.1 ; White Girls, Elementary, 16.2, High School, 27.7 ; Colored Boys, Elementary, 10.2, High School, 22.0 ; Col- ored Girls, Elementary, 8.9, High School, 23.9. Mixed Rela- tions Test II — ^White Boys, Elementary, 18.5, High School, 28.6; White Girls, Elementary, 17.0, High School, 33.7; Col- ored Boys, Elementary, 12.3, High School, 31.8 ; Colored Girls, Elementary, 10.2, High School, 29.0. Completion Test— White Soys, Elementary, 18.4, High School, 28.6; White Girls, Ele- COMPARISON OF WHITES AND NEGROES. 57 mentary, 16.1, High School, 33.1; Colored Boys, Elementary, 12.6, High School, 23.5 ; Colored Girls, Elementary, 15.6, High School, 25.4. The maze and cancellation tests did not bring out marked differences anywhere, as will appear, and the dif- ferences revealed by them in this comparison of the elementary and high schools were not considerable. These figures show that the white high school pupils ob- tained on the whole nearly twice the score that was reached by the elementary pupils of the same age; and that the col- ored high school pupils obtained on the whole more than twice the score that was received by the equally old elementary pupils. Thus is borne out the contention that the older pupils in the upper grades of the elementary school are the poorest of their age; the brightest of their age have passed beyond the elementary school. And it is reasonable to suppose that the same state of affairs exists in the upper high school grades. The able 18 year old students have gone out of the school and left behind their less able fellows. And these are inferior to pupils of normal high school ability at a lower age. It follows as a corollary of the above, that the youngest pupils represented in the graphs have scores which are too high to be truly representative of their ages. The lowest ages in any limited number of grades contain only the brightest pupils of those ages; the mediocre and the dull have not yet reached the grades in question. While the highest ages score too low, the lowest ages score too high. So it appeared that the youngest and the oldest pupils in the elementary grades tested in Richmond stood close together in their scores, and this was probably on account of the undue superiority of the young as much as on account of the undue inferiority of the old. The same is true of the high school taken alone. Ages 14 to 18, inclusive, were represented in the Richmond high schools in sufficient numbers for comparison. The scores received by the different ages of both white and colored pupils behaved in the same way; the scores of the white high school pupils follow: Mixed Relations Test I— Boys, 23.1, 25.2, 31.3, 25.6, 25.0; Girls, 30.3, 25.2, 25.6, 20.1, 16.6. Mixed Relations Test H— Boys, 29.0, 28.3, 36.8, 34.6, 29.6; Girls, 35.4, 32.0, 32.5, 28.0, 23.0. As in the case of the elementary school, the oldest and the youngest high school pupils do not stand far apart in their scores. And as the older pupils scored too low, so the younger pupils scored too high. ' This is evidenced by 58 THE PSYCHOLOGY OF THE NEGRO. the fact that when all of the pupils of their age, both elemen- tary and high school, are considered together, the resulting score, which is the true one, is much lower. Of course it is the combined score of both high school and elementary pupils of the same age that is represented in the tables and graphs. This suggests a matter to which attention should be called. The attempt is sometimes made to establish norms for vari- ous ages of pupils in psychological tests. The idea is to as- certain by a large number of experiments the performance that may be expected from a child of a certain age. But it is very evident that such attempts can only serve to mislead if they deal, say, with pupils in the higher elementary grades without taking the high school into consideration, or if they deal with the lower grades of the high school without taking the elementary school into consideration. In the former case, the resulting norms for the higher ages will be much too low ; in the latter case, for the same ages, they will be much too high. A fourteen or fifteen year old child in the school system is not a typical fourteen or fifteen year old child. He is typi- cal of a child of his age in high or in elementary school, and not in general. The only way in which valid age norms may be established is by testing the ages throughout their normal distributions among the grades. Or if only one school is tested, the elementary or the high school, it should at least be stated that such is the case, so that one may be able to make proper allowances for the findings reported. Even this has not always been done. Attention may also be called to the bearing of the compara- tives scores of the elementary and high school pupils of the same age upon a contention that was made in the preceding chapter. These scores show that the high school pupils are superior to the elementary school pupils of equal age to a considerably greater extent in the case of the negroes than of the whites. The high school negroes more than doubled the score of the elementary school negroes in each of the four comparisons in the mixed relations test, but the high school whites did not double the score of the elem.entary whites in any of the four comparisons ; in the completion test, the high school doubled the score of the elementary school only in the case of the white girls. This fits in with the other indications that the colored high school is a more closely selected group from the point of view of ability than is the white high school. COMPARISON OF WHITES AND NEGROES. 59 TABLE VII. Mixed Relations Test — Difference Between Scores op; White AND Colored Subjects Classified by Age and Sex (Minus signa indicate greater scores by the colored subjects.) Richmond. Ages 11 12 13 14 15 16 17 18 Av. P.E. Test I. Boys 3.3 6.3 6.5 4.3 7.2 7.3 3.6 5.0 5.4 .4 Girls 4.5 8.2 5.6 8.2 4.4 3.9 —6.2 —1.8 3.3 1.1 Test II. Boys 3.9 7.9 7.6 7.8 2.8 7.3 2.6 —2.5 4.7 .9 Girls * 4.8 11.8 9.7 8.3 6.5 3.8 —4.1 .3 5.1 1.2 Fredericksburg. Test I. Boys 4.9 .. 13.5 12.3 .. 16.0 11.7 1.4 Girls 8.8 12.0 13.3 —11.9 5.5 7.0 5.8 2.1 Test II. Boys 9.3 . . 13.1 13.1 . . 9.0 11.1 .8 Girls 5.8 18.0 6.6—10.2 4.6 5.0 4.9 1.8 Newport News. Test I. Boys 2.9 .. 11.9 7.4 2.7 Girls . . 15.4 11.7 10.3 8.7 11.5 .8 Test II. Boys 2.9 .. 13.8 8.3 3.2 Girls 21.5 19.7 15.2 7.9 16.1 1.9 TABLE VIIL Mixed Relations Test — Difference Between Scores of White and Colored Subjects Classified by Grades (Minus signs indicate greater scores by the colored subjects.) Richmond. Grades 5A 5B 6A 6B 7A 7B lA 2A 3A 4A Av. P.E. test 1 4.2 3.6 3.8 3.4 8.7 12.7 4.1 7,3 —6.3 .7 4.2 .& Test II 5.1 7.7 5.8 8.7 5.3 17.6 1.9 5.5 —3.6 .6 5.5 .9 Fredericksburg. Test 1 5.5 .. 10.3 .. 9.5 6.4 3.8 6.6 7.0 .7 Test II 8.5 .. 10.9 .. 11.9 4.3 2.6 4.2 7.1 1.6 Newport News. Test 1 15.1 15.1 .. Test II 17.8 .. .. 17.8 .. TABLE IX. Mixed Relations Test — Percentage of the Score of the White Obtained by the Colored Subjects, Classified by Age and Sex Richmond. Ages 11 12 13 14 15 16 17 18 Av. P.E. Test I. Boys 78 61 64 76 64 74 86 80 72.9 2.2 Girls 74 57 69 64 81 84 131 111 83.9 5.7 Test II. Boys 76 58 64 67 87 78 93 108 78.9 3.9 Girls 78 46 54 67 78 88 115 99 78.1 5.1 Fredericksburg. Test I. Boys 65 .. 50 47 .. 58 55.0 2.7 Girls 54 45 30 170 80 65 74.0 11.9 Test II. Boys 48 . . 60 59 . . 76 60.7 3.2 Girls 69 38 71 141 86 82 81.2 7.7 60 THE PSYCHOLOGY OF THE NEGRO. Newport News. Test I. Boys 82 .. 37 59.5 13.2 Girls 33 38 36 21 32.0 2.3 Test II. Boys 88 . . 34 61.0 16.2 Girls 31 18 31 34 28.5 2.3 TABLE X. Mixed Relations Test — Percentage of the Score op the White Obtained by the Colored Subjects, Classified by Grades Richmond. Grades 5A 5B 6A 6B 7A 7B lA 2A 3A 4A Av. P.E. Test I.. 68 74 78 80 52 49 84 74 127 97 78.3 3.9 Test II. 66 57 71 62 76 39 93 84 111 98 75.7 4.4 Test I, . 66 F] rederic 43 :ksbui 59 72 85 79 67.3 3.8 Test II. , . , , 55 50 . . 63 85 92 88 72.2 5.6 Newport News. Test I. . , , , , , , , , 31 , , , , , , , , . . 31.0 ^ ^ Test II. • • 29 • • .. •• . . 29.0 •• TABLE XI . Mixed '. Relations ' Iest- -Percentage of Colored Subjects Reaching or Exceeding the Average of the White, by Age and Sex AND BY Grades Richmond. Ages 11 12 13 14 15 16 17 18 Av. P.E. Test I. Boys . . ..33 14 21 37 15 25 33 17 ^ , .. 24.4 2.2 Girls .. .. 22 11 26 21 32 33 78 53 .. 34.5 4.5 Test II. Boys . . .. 33 15 16 28 47 21 56 67 , , .. 35.4 4.8 Girls .. .. 25 10 15 29 37 42 74 53 .. 35.6 4.8 Grades 5A 5B 6A 6B 7A 7B lA 2A 3A 4A Test I. . .. 21 20 37 22 13 11 34 29 87 56 33.0 4.2 Test II, ...18 19 24 18 32 62 26 88 52 33.9 5.2 To return to the comparative scores of negroes and whites in the mixed relations test. Tables 7-11 give the detailed comparisons. In every instance, as shown by the averages, the whites surpassed the negroes, and the probable errors are so small as to render the averages very reliable. In Test I, in Richmond, the colored boys obtained 72.9 per cent, of the score of the white boys; the colored girls obtained 83.9 per cent, of the score of the white girls; the colored grades ob- tained 78.3 per cent, of the score of the white grades. In Test II, the colored boys and girls obtained 78.9 and 78.1 per cent., respectively, of the score of the white boys and girls, and the colored grades scored 75.7 per cent, as high as did the white grades. The figures are very constant. We may conveniently COMPARISON OF WHITES AND NEGROES. 61 average the percentages for boys, girls and grades, and find that in Test I the colored pupils obtained 78.4 per cent, of the score of the whites, and that in Test II they obtained 77.6 per cent, of the white score. In Fredericksburg and Newport News the percentage of the score of the whites obtained by the negroes was considerably smaller than in Richmond. The figures showing the percentage of Richmond negroes reaching or exceeding the average of the whites are as follows : Test I— Boys, 24.4; Girls, 34.5; Grades, 33.0. Test II— Boys, 35.4 ; Girls, 35.6 ; Grades, 33.9. Averaging the percentages for boys, girls and grades, we find that in Test I 30.6 per cent, of the negroes reached or exceeded the white score, and that in Test II 35.0 per cent, of them reached or exceeded the score of the whites. It is apparent that this test reveals a considerable differ- ence between the two races. In both the first and second tests the difference is about equally marked, and it is approxi- mately the same for both boys and girls and for both ages and grades. Completion Test The actual scores in the Completion Test appear in Tables 12 and 13, classified by age and sex and by grades. These tables are the basis of Figures 7-9, and of Tables 14-18, in which the detailed comparisons are made. In this test, as in the mixed relations, though to a less degree, may be noted the relative superiority of the higher colored grades, and the fact that the higher ages, on the whole, do not show pro- gressively increasing scores as do the lower ages. TABLE XII. Completion Test — Scores by Age and Sex Richmond. 10 11 12 13 14 15 16 17 18 15.3 18.7 18.7 21.2 23.4 30.7 30.7 25.2 4.3 7.2 6.9 8.2 6.4 7.4 6.9 5.8 A.D. 4.6 3.1 3.0 3.7 6.3 5.2 5.0 6.3 3.8 19 Ages Boys — white Av A.D Boys— Col. Av A.D Girls — ^white Av A.D Girls— Col. Av 13.6 16.8 13.9 14.5 17.9 22.0 24.0 28.1 23.9 28.6 8.8 13.2 13.6 15.8 16.1 21.2 28.1 31.3 27.2 1.8 2.2 4.8 6.1 8.6 7.8 7.6 5.6 5.2 18.9 20.9 17.0 25.3 27.7 28.9 30.0 33.2 33.5 4.3 5.7 5.0 8.5 8.3 7.8 8.0 3.0 .5 .6 62 THE PSYCHOLOGY OF THE NEGRO. Fredericksburg. Ages 10 11 12 13 14 15 16 17 18 19 Boys — ^white Av . . 24.6 12.0 19.1 23.0 23.7 33.0 31.5 43.3 ^ ^ A.D .. 8.0 4.0 7.2 10.0 9.0 6.0 5.5 3.0 ^ ^ Boys — Col. Av 15.6 10.0 , , 22.0 17.2 , , 26.5 24.5 A.D 3.3 3.0 10.0 5.4 , , 2.5 6.5 Girls — ^white Av . . 24.4 16.5 25.5 22.5 28.8 39.2 35.0 ^ , A.D . . . • 6.0 6.1 6.0 4.1 7.1 5.8 0. • e Girls— Col. Av . . 15.3 15.0 22.6 25.0 24.2 11.0 28.0 A.D . . . . 4.5 3.2 4.6 8.0 8.0 4.0 13.0 Newport News. Boys — ^white Av . . 25.3 23.2 24.7 20.4 14.8 ^ ^ , , ^ ^ ^ , A.D .. 2.3 5.6 6.8 5.3 3.2 ^ , Boys— Col. Av • • • • 16.5 ^ ^ 14.7 ^ ^ ^ , , . • •- • • A.D 2.5 2.7 Girls — ^white Av . . 30.8 23.0 21.8 21.0 21.5 22.0 ^ ^ A.D .. 4.8 5.2 5.2 5.4 3.5 5.0 Girls— Col. Av . . 13.5 13.1 12.5 8.6 9.0 ^ ^ ^ ^ A.D • • . . 5.7 2.5 4.4 2.0 5.3 , , , . Grades TABLE XIIL Completion Test — Scores by Grades Richmond. 5A 5B 6A 6B 7A 7B lA 2A 2B 3A 4A White Av .... 13.3 15.1 17.6 22.5 19.4 25.6 27.0 30.5 . . 31.2 34.5 A.D ... 4.6 4.7 4.6 3.8 5.2 6.8 7.1 6.7 .. 5.9 6.0 Col. Av . ... 11.3 13.2 14.2 16.5 16.1 15.7 21.8 27.2 23.4 28.4 29.4 A.D ... 3.5 2.6 4.1 4.0 4.3 4.4 6.3 6.3 Fredericksburg. 4.5 4.8 6.6 White Av • • • • • . . 17.5 . . 21.3 . . 26.7 26.5 . . 36.0 36.1 A.D .. 7.0 .. 6.0 .. 7.8 6.4 . . 5.1 6.0 Col. Av • • • • > • . . 13.0 . . 19.2 . . 18.4 20.7 . . 24.2 30.1 A.D .. 3.5 .. 4.8 .. 3.5 5.0 Newport News. . . 9.7 6.6 White Av • • • • • . . . . 21.9 21.9 23.7 . . . . A.D • • • • • .. ..7.5 5.0 6.5 .. .. Col. Av .. 12.6 .. 13.3 A.D • . • • • .. 3.8 .. 4.8 The colored boys in Richmond obtained 78.5 per cent, of the score of the white boys. The colored girls obtained 80.0 per cent, of the score of the white girls. And the colored grades obtained 81.7 per cent, of the score of the white grades. The average for boys, girls and grades is 80.1. In Fredericks- burg and Newport News the negroes are more inferior to the whites than in Richmond, as was the case in the mixed rela- tions test. And the difference between the races is about the same for both boys and girls and for both ages and grades. COMPARISON OF WHITES AND NEGROES. 63 Fig. 7. Completion Test — Scores of White and Colored Boys- Richmond. Fig. 8. Completion Test. — Scores of White and Colored Girls- Richmond. Fig. 9. Completion Test— Scores of White and Colored Grades- Richmond. 64 THE PSYCHOLOGY OF THE NEGRO. In Richmond the percentage of colored boys reaching or exceeding the average of the whites was 32.5. The percentage of colored girls reaching or exceeding the average of the whites was 24.7. By grades, the percentage of colored sub- jects reaching or exceeding the average of the whites was 26.5. The average for boys, girls and grades is 27.9. TABLE XIV. Completion Test — Difference Between Scores op White and Col- ored Subjects Classified by Age and Sex (Minus signs indicate greater scores by the colored subjects.) Richmond. Ages 11 12 13 14 15 16 17 18 Av. P.E. Boys Girls Boys Girls Boys Girls 6.5 5.5 5.1 5.4 7.3 9.5 2.6 —6.1 4.5 .9 2.1 7.0 2.5 7.4 5.7 4.9 1.9 9.3 5.1 .7 16.8 10.8 2.3 15.0 24.0 9.1 2.6 ..6.2 .3 ..9.9 .3 7.0 2.5 7.4 5.7 4.9 Fredericksburg. 9.1 1.7 15.8 1.2 10.5 — .1 3.8 Newport News. 6.7 , , 5.7 9.5 8.7 8.5 12.9 TABLE XV. Completion Test — Difference Between Scores op White and Col- ored Subjects Classified by Grades (Minus signs indicate greater scores by the colored subjects.) Richmond. Grades 5A 6B 6A 6B 7A 7B 1A 2A 3A 4A Av. P.E. 2.0 1.9 3.4 6.0 3.3 9.9 5.2 3.3 2.8 Fredericksburg. 4.5 ..2.1 .. 8.3 5.8 11.8 Newport News. . . 8.6 5.1 4.3 6.0 6.4 .. 8.6 .5 .8 TABLE XVI. Completion Test— Percentage of the Score op the White Obtained BY THE Colored Subjects, Classified by Age and Sex Richmond. Ages 11 12 13 14 15 16 17 18 Av. P.E. Boys Girls 57 89 71 67 73 85 74 68 69 92 124 70 80 83 94 72 78.5 80.0 4.2 2.3 Boys Girls •• •• 52 92 Fredericksburg. 93 52 .. 61 58 100 87 62 31 64.5 71.7 6.9 7.3 Boys Girls •• 71 59 60 Newport News. 71 60 39 71.0 54.5 .0 3.4 COMPARISON OF WHITES AND NEGROES. 65 TABLE XVII. Completion Test — Percentage op the Score of the White Obtained BY THE Colored Subjects, Classified by Grades Richmond. Grades 5A 5B 6A 6B 7A 7B lA 2A 3A 4A Av. P.E. 85 87 81 73 83 62 81 89 91 85 81.7 1.5 Fredericksburg. .. .. 74 .. 90 .. 69 78 67 83 76.8 2.3 Newport News. 61 61.0 TABLE XVIII. Completion Test — Percentage of Colored Subjects Reaching ob Exceeding the Average of the White, by Age and Sex and by Grades Agres 11 12 13 Richmond. 14 15 16 17 18 Av. P.E. Boys Girls 33 14 9 26 41 35 25 21 17 19 24 56 48 83 .. 7 .. .. 32.5 . . 24.7 5.7 3.& Grades 5A 5B 6A 6B 7A 7B lA 2A 3A 4A 33 34 20 15 32 26 29 48 28 26.5 2.3 The racial differences are very similar in all respects to those brought out by the mixed relations test. They are equally constant in the two tests, the small probable errors render them equally valid, and they are of approximately equal amounts. This was to be expected, since the two tests deal in general with the same mental traits, and the results from the two serve to reinforce and establish the validity of each other. The results from the two tests and the three cities and in the various modes of comparison make it in- dubitable that in the important mental capacities measured whites are much superior to negroes. Just how much the whites are superior in the traits meas- ured it is impossible to say. For we do not know the zero point of the tests. Averaging the results from the mixed relations tests, I and II, and from the completion test, we find that in the tests themselves the negroes did 78.7 per cent, as well as the whites. But we cannot say more than this. We cannot say that the negroes have 78.7 per cent, of the ability of the whites, for we have no index of where the ability in question begins. 66 THE PSYCHOLOGY OF THE NEGRO. Maze Test The scores for the maze test are set forth in Tables 19 and 20. These tables, so far as they concern Richmond, are made plainer by Figures 10-12. An examination of these fig- ^lres shows that in every age for each sex, and in every grade •except one, the whites covered a greater distance and made more touches than did the negroes. This seems to be a racial difference for the subjects tested. It appears in the Freder- icksburg results as well as in those of Richmond. The negroes were more careful than the whites. The significance of this is difficult to tell. And it becomes especially difficult when it is considered in connection with the fact, which will appear, that the negroes were not more careful in the cancellation test. TABLE XIX. Maze Test — Scores by Age and Sex Richmond. Ages 10 11 12 13 14 15 16 17 18 19 Boys — white Touches Av. 39.7 48.6 48.2 45.5 57.2 63.3 , , , , , , A.D. 20.2 19.6 27.8 26.1 19.5 25.0 , , , , , , Distance Av. ^ , 94.4 103.2 99.7 99.4 113.8 123.0 , , . . , , A.D. 20.2 23.8 27.3 22.6 20.6 22.6 , , . . , , Boys— Col. Touches Av. 34.4 35.5 31.6 33.7 28.1 14.5 30.0 43.8 16.7 A.D. 14.8 23.0 19.7 21.8 22.5 7.7 24.2 31.5 11.7 Distance Av. 88.6 90.1 84.4 93.8 83.0 83.2 91.1 104.5 76.0 A.D. 12.6 23.9 22.6 28.0 20.2 9.0 23.0 21.5 30.0 Girls — ^white Touches Av. ^ , 54.0 60.0 41.0 44.3 59.6 , , , , , , , . A.D. 16.3 22.6 18.0 26.4 13.3 , , , , , , , , Distance Av. 113.7 117.0 110.8 103.8 119.3 ^ , , , , ^ , , A.D. 13.0 22.0 20.4 25.3 13.7 ^ ^ , , ^ , Girls— Col. Touches Av. 34.6 21.0 27.2 28.9 35.8 39.0 35.0 21.5 19.2 16.6 A.D. 16.0 15.2 15.9 19.4 22.7 25.4 25.3 16.7 10.5 10.0 Distance Av. 93.6 80.1 84.8 86.2 97.3 101.4 99.0 89.9 86.0 70.6 A.D. 6.0 27.1 23.7 22.9 24.3 17.5 25.5 18.9 17.5 20.6 Fredericksburg Boys — white Touches Av. , , 32.4 30.7 17.0 25.3 38.5 ^ ^ A.D. 21.0 7.2 9.0 15.3 8.3 , ^ Distance Av. , , , . , , 111.2 112.7 88.6 107.0 111.0 , , A.D. , ^ , , , , 19.1 20.2 12.6 23.6 9.3 ^ ^ Boys — Col. Touches Av. 26.6 22.5 ^ ^ 36.6 41.0 ^ , 14.0 34.0 A.D. ^ ^ , , 12.6 19.5 , , 6.6 36.8 , , 6.0 24.5 Distance Av. ^ ^ , , 76.3 70.0 , , 95.0 80.8 , . 69.5 97.5 A.D. ^ ^ 13.6 26.0 , , 14.6 34.0 ^ , 2.5 32.5 Girls — white Touches Av. ^ ^ ^ ^ 13.6 24.2 11.0 17.6 8.0 ^ ^ A.D. , . . , , . 4.6 15.2 8.0 9.4 4.0 COMPARISON OF WHITES AND NEGROES. 67 Ages 10 11 12 13 14 15 16 17 18 19 Distance Av. 84.6 7.3 94.0 13.5 85.0 20.3 97.4 5.6 68.0 1.0 A.D. ][ Girls— Col. Touches Av. . . 13.0 11.3 6.3 25.5 10.3 22.0 54.5 A.D. . . 9.5 6.7 4.3 6.5 7.7 10.0 44.5 Distance Av. . . 68.7 58.5 58.6 76.0 65.8 84.0 99.5 A.D. . . 21.2 14.3 14.0 3.0 8.8 40.0 Newport News. Boys— Col. Touches Av. .. 5.0 9.5 , , , , , , ^ , A.D. .. 3.0 3.0 ^ , ^ , ^ , Distance Av. . . 56.5 63.5 ^ ^ ^ ^ , ^ A.D. .. 7.5 11.0 ^ , ^ , ^ , Girls— Col. Touches Av. . . 17.4 10.6 7.0 6.0 10.3 , , , , , , A.D. . . 20.7 6.6 7.2 4.0 5.0 - . Distance Av. . . 72.2 54.0 59.2 59.3 61.6 . , , , , , A.D. . . 23.7 16.0 15.8 6.3 14.0 •• •• •• TABLE XX. Maze Test — Scores by Grades Richmond. Grades 5A 5B 6A 6B 7A 7B lA 2A 2B 3A 4A White Touches Av. 53.6 46.9 41.7 38.6 40.2 61.1 A.D. 22.5 23.4 23.1 22.0 24.8 17.7 Distance Av. 106.5 106.6 95.5 97.0 103.0 123.7 A.D. 21.3 24.0 25.3 23.2 22.9 13.7 Colored Touches Av. 38.3 26.0 21.7 18.6 46.1 46.7 31. 2i 34.3 33.9 20.5 16.8 A.D. 20.6 18.3 13.4 14.7 15.9 26.9 21.3 24.2 23.7 15.2 9.0 Distance Av. 100.4 85.3 70.7 69.5 111.7 103.3 94.6 98.7 101.0 84.2 81.8 A.D. 23.1 21.7 14.5 20.2 15.2 24.5 Fredericksburg. 19.3 18.0 25.0 17.9 20.0 White Touches Av. , , , , 26.1 .. 24.6 .. 31.6 13.5 19.9 25.8 A.D. , , , , 16.5 .. 15.7 .. 17.7 8.4 11.0 15.3 Distance Av. , , 94.2 .. 92.2 .. 104.6 87.3 97.9 102.7 A.D. , , 24.4 .. 20.4 .. 22.3 13.0 16.7 17.8 Colored Touches Av. 19.2 .. 16.5 .. 10.5 21.8 24.5 31.7 A.D. 15.4 .. 11.1 5.7 16.2 14.5 25.7 Distance Av. 67.1 .. 72.1 61.2 83.0 78.3 90.7 A.D. 18.8 .. 20.4 .. Newport News. 14.2 16.0 14.8 28.0 Colored Touches Av. , , 13.1 .. 6.8 .. , , ^ ^ A.D. 11.5 .. 5.0 .. Distance Av. , , 67.5 .. 55.6 .. , ^ . ' * ] A.D. « , 16.6 .. 13.4 .. , . , , . , It is not apparent that there is any general increase or decrease in either the distance covered or the number of touches, or in the ratio between the two, as the age of the pupils advances. The difference in ages, indeed, may fairly be disregarded. 68 THE PSYCHOLOGY OF THE NEGRO. Fig. 10. Maze Test — Scores of White and Colored Boys — Richmond.* *The total height of the columns indicates the score for distance; the height of the cross-lines indicates the score for touches. COMPARISON OF WHITES AND NEGROES. 69 Fig. 11. Maze Test — Scores of White and Colored Girls — Richmond. 70 THE PSYCHOLOGY OF THE NEGRO. SCO HE • 110 GRADE Fig. 12. Maze Test — Score of White and Colored Grades — Richmond. COMPARISON OF WHITES AND NEGROES. 71 When we attempt to decide which race was superior in this test we encounter the difficulty in evaluating speed and ac- curacy that was discussed in the preceding chapter. Tables 21-25 give the detailed comparisons for both speed and accu- racy — distance and touches. The Fredericksburg compari- sons, on the whole, are fairly sure, although the probable errors are large. The tables show that in Fredericksburg the colored subjects exceeded the white subjects in number of touches made, but were inferior to them in distance covered, and that this was true for both sexes and for the grades. The negroes attained less speed and were at the same time more inaccurate. Consequently they were inferior to the whites. But in Richmond a conclusion cannot be drawn so readily. Here the colored subjects were slower, as in Fredericksburg, but, unlike those in Fredericksburg, they were also more accu- rate than the whites. The colored boys covered 86.6 per cent, of the distance covered by the white boys and made 69.6 per cent, as many touches. The colored girls scored 79.8 per cent, of the distance of the white girls and 60.2 per cent, as many touches as the white girls. The scores for the grades show that the colored pupils scored for distance and touches, re- spectively, 85.2 per cent, and 70.0 per cent, as high as did the white pupils. These figures are fairly uniform for the differ- ent classifications of the subjects, and their probable errors are sufficiently small for reliability. If we average the per- centages given for boys, girls and grades, we find that the negroes covered 84 per cent, as great distance as the whites and made 67 per cent, as many errors. TABLE XXI. Maze Test — Difference Between Scores of White and Col- ored Subjects Classified by Age and Sex (Minus signs indicate greater scores by the colored subects.) Richmond. Ages 11 12 13 14 15 16 17 18 Av. P.E. Boys Touches 5.3 13.1 16.6 11.8 29.1 , ^ 15.2 2.3 Distance 5.8 13.1 15.3 5.6 30.8 , , 14.1 2.7 Girls Touches 33.0 32.8 12.1 8.5 20.6 21.4 3.5 Distance 33.6 32.2 24.6 6.5 17.9 . , 23.0 3.3 Fredericksburg Boys Touches , , —5.9- -24.0 24.5 —1.8 8.6 Distance , , 17.7 7.8 41.5 22.3 6.3 Girls Touches ^ , 2.3 17.9 — -14.5 7.3- -14.0 —.2 4.3 Distance 26.1 35.4 9.0 31.6- -16.0 17.2 6.3 72 THE PSYCHOLOGY OF THE NEGRO. TABLE XXII. Maze Test — Difference Between Scores of White and Ck>L08ED Subjects Classified by Grades (Minus signs indicate greater scores by the colored subjects.) Richmond. Grades 5A 5B 6A 6B 7A 7B lA 2A 3A 4A Av. P.E Touches 15.3 20.9 20.0 20.0 —5.9 14.4 14.1 2.3 Distance 6.1 21.3 24.8 27.5 ^8.7 20.4 15.2 3.8 Fredericksburg. Touches .. ..6.9 .. 8.1 .. 21.1—8.3—4.6—5.9 2.9 3.2 Distance . . . . 27.1 . . 20.1 . . 43.4 4.3 19.6 12.0 21.1 3.3 TABLE XXIII. Maze Test — Percentage of the Score of the White Obtained by the Colored Subjects, Classified by Age and Sex Ages Boys Touches Distance Girls Touches Distance Boys Touches Distance Girls Touches Distance 11 12 87 73 94 87 Richmond. 13 14 15 65 85 74 94 49 73 39 45 70 81 66 70 72 78 94 85 Fredericksburg. 16 17 18 Av. P.E. 69.6 3.8 86.6 2.2 119 241 84 91 60.2 5.7 79.8 2.9 36 132.0 36.1 63 79.3 5.4 84 25 232 59 275 135.0 36.2 89 62 89 67 124 86.2 6.4 TABLE XXIV. Maze Test — Percentage of the Score of the White Obtained by the Colored Subjects, Classified by Grades Richmond. 5A 5B 6A 6B 7A 7B lA 2A 3A 4A Av. P.E. Grades Touches Distance Touches Distance 70.0 6.3 3.5 72 56 52 49 115 76 94 80 74 72 108 83 85.2 Fredericksburg. . . . . 73 . . 68 . . 34 164 123 123 97.5 13.5 .. .. 71 .. 78 .. 59 95 80 88 78.5 3.2 TABLE XXV. Maze Test — Percentage of Colored Subjects Reaching or Exceed- ing the Average of the White, by Age and Sex and by Grades Richmond. Ages 11 12 13 14 15 Av. P.E. Boys Touches 40 20 22 29 14 25.0 2.9 Distance 40 30 39 35 14 31.6 2.9 COMPARISON OF WHITES AND NEGROES. 73 Ages Girls Touches Distance 11 25 12 10 20 13 35 29 14 30 41 15 22 19 •• Av. 19.4 26.8 P.E. 4.5 2.5 Grades 5A 5B 6A 6B 7A 7B Touches Distance 26 41 19 22 14 11 14 14 68 68 35 35 29.3 31.8 5.2 6.6 The percentages of colored subjects reaching or exceeding the average of the white sho[vv' the same relative standing for the two races in touches and distance traversed as is evident from the preceding figures. These percentages are as follows: Boys — Touches, 25.0, Distance, 31.6; Girls — Touches, 19.4, Distance, 26.8; Grades — Touches, 29.3, Distance, 31.8. If we average the percentages for boys, girls and grades, as before, we find that 25 per cent, of the negroes reached the average of the whites in touches, and that 30 per cent, of them reached the average of the whites in distance. Do these facts indicate that the colored subjects were superior, inferior or equal to the white? A reference to the graphs will make it evident that a large score for distance implies a relatively higher score for touches than does a small score for distance. The ratio of touches to distance becomes greater as distance increases. And this change in the ratio takes place with marked constancy throughout the graphs. It is normal for the smaller of two scores for distance to have a relatively smaller score for errors. The question to be set- tled in this comparison is whether a distance 84 per cent, as great as another distance normally implies 67 per cent, as many touches for the former distance as for the latter. If so, it is apparent that there is no diflference between the ability of the white and the colored subjects in this test. And the following considerations indicate that such is the case. Since the score of the negroes for distance is 84 per cent, of that of the whites, 100 may be taken to represent the actual distance score of the whites and 84 to represent the actual distance score of the negroes. Now a reference to the graphs shows that colored grades 5A and 7B, and colored girls aged 14 and 15, scored approximately 100 for distance. And the average number of touches made by these four groups of col- ored pupils was 40, the variation being slight. A further ref- erence to the graphs shows that colored grade 5B, colored girls aged 11, 12 and 13, and colored boys aged 11, 13 and 15, 74 THE PSYCHOLOGY OF THE NEGRO. scored approximately 84 for distance. And the average num- ber of touches made by these seven groups of colored pupils was 28, the variation again being slight. Here we have two sets of colored pupils scoring 100 and 84, respectively, for dis- tance, and 40 and 28 for errors. Twenty-eight is 70 per cent, of 40, and 84 is 84 per cent, of 100. Thus we have the slower set of colored pupils attaining 84 per cent, as great distance as the faster set, and making 70 per cent, as many errors. But these are approximately the figures which represent the difference between the whites and the negroes. So it seems that we can confidently conclude that there is no racial differ- ence in ability revealed by the maze test. Wherever the ne- groes work as rapidly as the whites, they make approximately the same number of errors. The graphs show no records of white subjects with as small scores as 84 for distance, and we therefore cannot ascertain the actual number of errors made by whites who worked as slowly as the assumed average of the negroes. But there is no reason for believing there would have been any appreciable racial difference if slow white records had occurred. When we compare the white groups who attained a distance of 100, the assumed average of the whites, with the colored groups who worked rapidly enough to attain this distance, we find the number of errors made by the two groups to be about the same. White grades 6A, 6B and 7A, white girls aged 14, and white boys aged 12, 13 and 14, scored approximately 100 for distance. And the average number of errors of these seven groups, the variation being small, was 44. In the preceding paragraph it was pointed out that the four colored groups who worked at this speed scored an average of 40 errors. The difference between the two sets of pupils, white and colored, is thus inconsiderable. Finally, the reader may be referred to the white and col- ored columns in the graphs themselves. An inspection shows that wherever the column for the whites approaches that for the negroes in total height, representing distance, the cross- lines on the two columns, representing errors, correspondingly approach equality. And within the colored columns them- selves, as within the white columns, greater total height means not only an absolutely but a relatively greater height for the cross-lines. On the whole, while the whites were superior to the negroes COMPARISON OF WHITES AND NEGROES. 75 in Fredericksburg in the motor ability tested, they were not superior in Richmond. Greater weight shguld be attached to the Richmond results on account of the larger number of sub- jects tested there. 1 Cancellation Test The records for the cancellation test are set forth in Ta- bles 26 and 27, and in Figures 13-15. A fairly regular in- crease in the scores with age may be noted. TABLE XXVI. Cancellation Test — Scores by Age and Sex Ages Boys — white Omissions Av. A.D. Cancellations Av. A.D. Boys— Col. Omissions Av. A.D. Cancellations Av. A.D. Girls — white Omissions Av. A.D. Cancellations Av. A.D. Girls— Col. Omissions Av. A.D. Cancellations Av. A.D. Boys — white Omissions Av. A.D. Cancellations Av. A.D. Boys— Col. Omissions Av. A.D. Cancellations Av. A.D. Girls — ^white Omissions Av. A.D. Cancellations Av. A.D. Girls— Col. Omissions Av. A.D. Cancellations Av. A.D. 10 Richmond. 11 12 13 14 15 16 17 18 19 1.0 3.2 2.8 3.2 1.6 2.0 1.0 2.2 2.4 2.3 1.1 .6 50.3 62.7 57.7 59.1 62.7 65.3 6.8 11.1 12.3 8.1 8.2 3.0 4.5 1.7 2.0 2.6 5.1 2.5 2.0 2.0 4.5 5.5 1.8 1.8 2.6 4.0 1.5 2.0 1.0 1.5 43.6 58.6 52.6 66.2 68.6 74.4 69.8 93.8 93.7 11.6 8.6 7.1 8.1 12.8 13.7 6.4 4.5 2.2 .8 2.5 1.8 2.9 3.5 .8 2.2 1.6 1.9 1.0 53.0 61.4 63.0 71.2 64.5 7.5 8.8 7.4 14.7 7.5 2.0 1.6 2.0 1.7 60.3 59.4 2.3 11.3 3.3 2.3 65.3 11.4 3.0 2.6 69.7 9.9 2.7 1.8 74.4 13.1 Fredericksburg. 1.3 1.0 53.3 5.0 3.6 2.4 63.9 11.4 1.5 3.1 1.5 l.S 45.0 54.7 3.0 6.7 3.3 1.0 1.6 0. 58.6 55.5 4.6 2.5 1.5 1.5 2.2 1.1 1.1 2.0 56.3 61.0 67.7 8.0 10.3 13.2 3.4 2.9 81.2 13.0 3.7 3.8 66.8 10.2 2.9 2.6 84.9 11.2 1.8 1.4 89.8 8.9 3.4 2.1 87.0 8.8 0. 0. 87.6 10.6 2.3 2.0 2.3 2.3 2.0 1.6 72.1 72.8 78.3 17.0 16.5 7.0 9.6 2.2 8.0 1.0 70.6 76.2 17.3 13.4 1.5 5.5 1.5 4.5 90.0 81.2 5.0 2.7 3.3 3.0 73.3 6.3 1.8 1.0 4.0 1.5 .8 4.0 72.0 75.2 74.5 13.0 12.6 4.5 1.6 2.0 .6 .6 1.8 .6 68.8 63.9 66.0 10.6 10.3 7.3 5.5 .5 94.5 .5 3.2 2.0 2.5 2.0 0. 1.5 84.7 81.5 70.5 9.7 16.5 25.5 76 THE PSYCHOLOGY OF THE NEGRO. Ages Boys — ^white Omissions Av. A.D. Cancellations Av. A.D. Boys— Col. Omissions Av. A.D. Cancellations Av. A.D. Girls — ^white Omissions Av. A.D. Cancellations Av. A.D. Girls— Col. Omissions Av. A.D. Cancellations Av. A.D. 10 Newport News. 11 12 13 14 15 16 17 18 19 .3 1.5 1.5 1.6 4.0 .3 .9 1.6 1.1 3.0 67.0 53.4 59.8 63.1 56.7 6.0 8.0 0. 0. 49.0 0. 8.7 8.1 2.2 1.2 69.5 7.0 20.2 .4 2.6 2.4 2.8 3.7 2.0 .4 2.6 2.4 2.1 2.7 2.0 52.4 60.7 57.4 71.0 63.0 72.5 6.0 9.8 11.3 11.7 9.5 15.5 3.7 1.4 1.6 1.6 9.0 3.0 .8 1.5 1.3 6.6 63.4 57.4 62.6 56.3 80.3 13.5 6.8 7.3 7.0 8.3 TABLE XXVII. Cancellation Test — Scores by Grades Grades White Omissions Av. A.D. Cancellations Av. A.D. Colored Omissions Av. A.D. Cancellations Av. A.D. White Omissions Av. A.D. Cancellations Av. A.D. Colored Omissions Av. A.D. Cancellations Av. A.D. White Omissions Av. A.D. Cancellations Av. A.D. Colored Omissions Av. A.D. Cancellations Av. A.D. Richmond. 5A 5B 6A 6B 7A 7B lA 2A 2B 3A 4A 2.6 3.0 1.5 1.7 2.5 2.1 1.8 2.9 1.1 1.5 2.1 1.4 61.6 55.2 54.4 65.5 62.8 66.8 9.9 7.4 11.0 12.6 8.2 10.0 4.0 3.4 1.4 2.8 3.7 5.5 2.9 2.2 2.6 2.4 2.1 3.5 3.3 1.3 2.4 2.8 4.5 2.3 1.9 1.9 1.9 1.7 60.2 58.3 63.1 65.6 81.6 74.5 77.4 82.4 89.1 86.4 87.1 10.9 9.5 9.0 8.6 11.2 11.2 12.7 14.0 8.7 9.1 10.0 Fredericksburg. . 2.3 . 2.2 . 53.0 6.5 .. 3.3 .. 2.8 .. 62.3 .. 9.9 . . 2.0 1.9 . . 1.8 1.2 . . 67.0 74.9 .. 7.8 9.9 . . 2.6 2.3 . . 2.0 2.0 . . 72.1 74.9 . . 14.7 13.8 3.0 2.9 57.7 6.5 Newpor .. 1.7 .. 1.0 .. 79.5 .. 7.7 t News. . . 2.1 4.3 . . 1.1 2.5 . . 76.4 88.3 .. 15.3 8.5 .. 3.3 2.2 . . 1.8 2.0 . . 82.0 78.1 . . 9.5 12.5 2.0 . 2.0 58.5 9.4 2.2 2.5 1.5 2.1 2.5 1.2 58.4 62.7 63.1 11.3 9.0 13.0 . . 4.6 . . . . 3.4 . . .. 70.0 .. COMPARISON OF WHITES AND NEGROES. 77 Fig. 13. Cancellation Test — Scores of White and Colored Boys — Richmond.* *The total height of the columns indicates the score for cancellations; the height of the cross-lines indicates the score for omissions. A reference to the graphs will show that there is no con- stant relation between the omissions and the cancellations, and that in all cases the omissions were very few. Only a small number of pupils, less than one-half, made any omissions at all. This accounts for the large deviations of the scores for omissions as shown in the tables mentioned, and for the very erratic behavior of these scores in the tables of ccompari- son. Tables 28-32. On the whole, the negroes seem to have made 78 THE PSYCHOLOGY OF THE NEGRO. Fig. 14. Cancellation Test — Scores of White and Colored Girls — Richmond. slightly more omissions than the whites, but the size of the probable errors indicates that this may have been a chance occurence. Certainly the negro girls, who, as will appear, are definitely superior to the white girls in cancellations, do not show as great excess in omissions as do the negro boys, and the negro boys are not superior to the white boys in can- cellations at all. The omissions, therefore, will be disregarded COMPARISON OF WHITES AND NEGROES. 79 SCORE 70 Fig. 15. Richmond. Cancellation Test — Scores of White and Colored Grades — in the following comparisons. It may be mentioned that the peculiarity noted above in the performance of the maze test by the two races, wherein the negroes made uniformly smaller scores and thus seemed to be more careful than the whites, does not occur in the test for cancellation. Another matter that may be remarked is that the large deviations of the scores for omission is in decided contrast with the small deviations of the cancellation scores, when the two sets of deviations are compared with the relative size of the scores themselves. That is, the so-called coefficients of variation are larger for omissions than for cancellations. A similar situation appeared in the maze test. Here the devia- 80 THE PSYCHOLOGY OF THE NEGRO. tions of the scores for touches were as large as those of the scores for distance, while the former scores were not over one-third as large as the latter. Now both omissions and touches are measures of accuracy and both cancellations and distance are measures of speed; the two tests measure the same phenomena, one in the field of movement and the other in the field of perception and reaction. And in both tests there is greater uniformity in speed of performance than in accuracy. TABLE XXVIII. Cancellation Test — Difference Between Scores op White and Colored Subjects Classified by Age and Sex (Minus signs indicate greater scores by the colored subjects.) Richmond. Ages 11 12 Boys Omissions — 3.5 1.5 Can'l't'ns 6.7 4.1 Girls Omissions — .8 — .8 13 14 15 16 17 18 Av. P.E. .8 .6 — 3.5 5.1 — 7.1 — 5.9 •• •• - .8 .8 .6 2.2 — 1.2 .2 .1 — 6.7 — 3.2 —16.7 •• •• - .5 .2 — 7.4 1.4 Fredericksburg. 2.1 . . — 5.9 — .8 . . — 3.8 - — .1 .2 2.7- — 7.8 3.8 7.3- .1 - 4.1 - 3.7 — -22.5 — 2.2 9.5- .8. -11.7 - 2.0- - 7.0- — .7 1.0 - 5.1 1.4 - .2 .6 — 5.9 2.7 Newport News. . . — .6 . . — 6.4 •• , , , . - .4 .6 - 1.0 3.2 1.0 1.2 2.1 0. 8.4 6.7 •• •• •• .8 .4 3.1 1.8 Boys Omissions Can'l't'ns Girls Omissions Can'l't'ns Boys Omissions . . 1.5 Canlt'ns . . 4.4 Girls Omissions . . — 1.1 Can'l't'ns . . — 2.7 TABLE XXIX. Cancellation Test — Difference Between Scores of White and Colored Subjects Classified by Grades (Minus signs indicate greater scores by the colored subjects.) Richmond. Grades 5A 5B Omissions —1.4 — .4 Cancellations 1.4 —3.1 - Omissions 6A 6B .1 —1.1 -8.7 — .1 7A - 1.2 7B -3.4 lA 2A 3A -18.8 —7.7 Fredericksburg. Cancellations .. —4.7 Omissions Cancellations 1.6 — .1 — 2.4 4A Av.f P.E .. —1.2 .3 .. —6.2 2.0 .1 — .4 .3 -17.2 . . —9.4 - Newport News. — 2.1 — 7.3 -13.4 —9.9 —3.2 —9.6 1.4 —2.1 —7.3 COMPARISON OF WHITES AND NEGROES. 81 TABLE XXX. Cancellation Test — Percentage of the Score of the White Ob- tained BY THE Colored Subjects, Classified by Age and Sex Richmond. Ages 11 12 13 14 15 16 17 18 Av. PJl. Boys Omissions 450 53 71 81 319 194.8 57.2 Cancellations 87 93 91 112 109 98.4 3.6 CtItIs Omissions 200 132 167 93 97 137.8 14.1 Cancellations 112 106 111 105 126 112.0 2.1 Fredericksburg. Boys Omissions .. .. 32 ..259 96 .. 65 113.0 30.1 Cancellations .. ..102 ..106 106 .. 115 107.2 1.5 Girls Omissions .. ..107 91 18 305 320 50 148.5 38.1 Cancellations .. ..113 94 90 131 113 109 108.3 3.8 Newport News. Boys Omissions .. ..137 68.5 4«.i.7 Cancellations .. 92 ..110 101.0 5.4 Omissions ..141 58 57 43 74.7 13.8 Cancellations .. 104 100 88 89 95.2 2.8 TABLE XXXI. Cancellation Test — Percentage of the Score of the White Ob- tained BY THE Colored Subjects, Classified by Grades Richmond. 5A 5B 6A 6B 7A 7B lA 2A 3A 4A Av. P.E. Grades Omissions Cancellations Omissions Cancellations Omissions Cancellations 154 113 93 165 148 262 98 106 116 100 130 111 Fredericksburg. .. ..130 .. 52 .. .. ..109 ..128 .. Newport News. 184 .. 155.8 13.2 110.2 3.0 105 226 127 96 122.7 13.2 114 118 114 104 114.5 1.9 184.0 .. 112 112.0 TABLE XXXIL Cancellation Test — Percentage of Colored Subjects Reaching or Exceeding the Average of the White , by Age and Sex and BY Grades Ages Boys Cancellations Girls Cancellations Grades Cancellations Richmomj . 11 12 13 14 15 Av. P.E. 33 31 37 78 60 47.8 6.4 67 59 74 60 81 68.2 2.8 5A 5B 6A 6B 7A 7B 49 63 80 55 89 67 67.2 3.8 82 THE PSYCHOLOGY OF THE NEGRO. In explanation of this it may be suggested that individuals differ in their attitude toward the tests as well as in their ability to perform them. Of course those who emphasize speed will neglect accuracy and those who emphasize accuracy will neglect speed. These matters will normally adjust them- selves among a large number of subjects, and the final scores for accuracy and speed will not be affected. But with refer- ence to approximating the average score it may be that speed of reaction will more readily take care of itself than will accuracy in a normal performance of a test. At the start, subjects are set to attempt a performance that will be in large measure speedy, and at least a reasonable degree of speed is put forth by all. Accuracy, however, requires more constant care on the part of the performer ; neglect of accuracy is not so apparent to him while he is reacting — he may not notice the omitted letter or the touch. It is more difficult to gauge and to control accuracy than speed. And so, unless there is a very careful adjustment of attitude toward accuracy by all subjects, it will vary considerably, while speed will vary to a less degree. Consequently a greater deviation for accuracy than for speed is to be expected in tests involving both factors. The tables which make the comparisons between whites and negroes in cancellation show the following facts : In Rich- mond, the colored boys obtained 98.4 per cent, of the score of the white boys. The colored girls obtained 112.0 per cent, of the score of the white girls. And when classified by grades — the grades contain a larger proportion of girls than of boys — the colored pupils obtained 110.2 per cent, of the score of the white pupils. This inferiority of colored to white boys may be disregarded, since it falls within the range of the probable error. It may be said that the boys of the two races did about equally well. But the colored girls and the colored grades are clearly superior to the white girls and grades, the probable errors in these cases being sufficiently small for reliability. In Fredericksburg, the colored boys, girls and grades are all definitely superior to the white. In Newport News, a very slight superiority is shown by the colored boys, but this may be discounted as was the inferiority of the colored boys in Richmond. The colored girls seem to be inferior here, but this is not well established when considered in connection with the probable error. The grade comparison is considerably COMPARISON OF WHITES AND NEGROES. 83 in favor of the colored pupils, but it is based upon figures from only one grade of each race. The percentages of colored subjects reaching or exceeding the average of the white, in Richmond, are as follows: Boys, 47.8; Girls, 68.2; Grades, 67.2. As in the former comparison for Richmond, the size of the probable errors renders it un- safe to say there was any real difference between the boys of the two races; but the colored girls and grades are un- doubtedly superior. Taken all together, the figures show that the colored girls are superior to the white girls in the traits measured by this test, and that the colored boys are not appreciably inferior to the white boys. This sex difference between the races is in- teresting, but no satisfactory explanation of it suggests itself. It is confirmed by Pyle ('15), in the study previously men- tioned. From the figures which he gives it may be calculated that in the "A Test" — the cancellation test used here — the col- ored boys tested by him obtained 98.4 per cent, of the score of the white boys and the colored girls obtained 108.2 per cent, of the score of the white girls. These are almost identical with the figures found herein. In this connection it may be said that no other constant or reliable racial sex difference in ability is indicated by this study. The scores of the boys of both races were generally, though not always, slightly lower than the scores of the girls, as is usually the case in psychological tests. And on the whole there is a slightly smaller difference between the white and negro girls than between the white and negro boys, but this is not at all invariably true. Pyle states, as quoted in Chapter I, that in the tests which he employed the girls of the two races stood nearer together than did the boys. CHAPTER IV COMPARISON OF SUB-CLASSES OF NEGROES The terms negro and colored have been used interchange- ably in this study, (except in the tables and the graphs), and they are generally so used. But it is obvious that such usage is justified only by convenience and not by fact. The so-called negroes of the United States are negroes only in part ; in large measure they are people of mixed blood, and the intermixture has been almost entirely with the white race. There are all degrees of intermixture, ranging from almost pure negro to almost pure "white. "Mulatto" is the term generally em- ployed to describe persons of this mixed stock. But here again the usage is justified only by convenience, since a mulatto, strictly speaking, is the offspring of the union of a pure white and a pure negro. The terms quadroon, meaning the child of a mulatto and a white, and consequently three parts white and one part negro ; and octoroon, meaning a child of a white and a quadroon, seven parts white and one part negro, are fairly common. But there is no recognized term to describe the racial status of the offspring of a pure negro and a mulatto, who is three parts negro and one part white. "Sambo" has been suggested for this purpose, but the popular connotation of the word is not such as to convey the desired meaning. And there are no widely accepted terms to describe other degrees of race admixture. The need for them has not been felt. The relative number of pure negroes and persons of part negro blood in this country is not accurately known. The Federal Census has made an attempt to determine this, and its figures are the most reliable that we have. The negroes were classified as black and mulatto. But as stated in the census report ('10, p. 79) : "Considerable uncertainty neces- sarily attaches to this classification, however, since the ac- 84 COMPARISON OF SUB-CLASSES OF NEGROES. 85 curacy of the distinction made depends largely upon the judg- ment and care of the enumerators. Moreover, the fact that the definition of the term 'mulatto' adopted at different cen- suses has not been entirely uniform may affect the compara- bility of the figures in some degree. In 1870, as in 1910, however, the term was applied to all persons having any per- ceptible trace of negro blood, excepting, of course, negroes of pure blood." The census shows that in 1910, 79 per cent, of the negro population of the country was black and 21 per cent, mulatto— roughly speaking, there were about eight million blacks and two million mulattoes. In Virginia, the negro pop- ulation was 67 per cent, black and 33 per cent, mulatto. In Richmond, the percentages of blacks and mulattoes were 60 and 40, respectively. The percentage of mulattoes is generally higher in the cities than in the rural districts. In Virginia, the five cities with a population of 25,000 or over have an average of 43 per cent, of mulattoes in their negro populations. This figure, when compared with the mulatto population of the state, is fairly typical of the larger proportion of mulattoes in the cities of the country at large. The percentage of mulattoes is also in general considerably higher in the North than in the South. The states with the largest negro populations have the largest percentage of pure negroes, though there are exceptions to this relation. South Carolina, Alabama and Georgia, for example, have negro populations of 55.2, 42.5 and 45.1 per cent., respectively. And the percentages of mulattoes among the negroes of these states are 16.1, 16.7 and 17.3, respectively. In Illinois, Pennsylvania and Massachusetts, the negroes are 1.9, 2.5 and 1.1 per cent, of the total populations. And in these states the percentages of mulattoes are 33.8, 19.2 and 36.7. States with negro popula- tions intermediate between the two groups mentioned are Kentucky, Tennessee and Maryland, which have in their re- spective populations 11.4, 21.7 and 17.9 per cent, of negroes. And here the percentages of mulattoes are respectively 25.2, 25.1 and 18.6. On the whole, there are larger percentages of mulattoes among the negroes in states with small negro populations, and there are larger percentages in the cities than in the country districts. What these facts indicate it is difficult to say. They may mean that mulattoes are more attracted than are pure negroes 86 THE PSYCHOLOGY OF THE NEGRO. by the reputed advantages held out to colored people in the North, and by the reputed opportunities of city life. If so, it would seem that mulattoes have greater capacity for perceiving opportunity and greater ambition to take advantage of it. Or color distinctions among the negroes themselves may influence those of lighter skin to seek to relieve the cir- cumpressure of the bulk of their race by Northern and urban migration. Or the non-agricultural industries in which ne- groes are engaged in the North and in the cities may possibly select mulattoes in preference to blacks, either because of their lighter color or because of their greater ability, or for both reasons. Or it may be that there is greater race inter- mixture in the North and in the cities than in the rural dis- tricts and the South. All four of these factors may operate. The census shows that there has been taking place a migra- tion of negroes from the country to the city, and from the South to the North — ^particularly to the cities of the North. And it is in these centers of migration that the proportion of mulattoes is highest. Several observers, (see Stone ,'08, p. 401, ff.), have reported the apparent growth of a class dis- tinction between mulattoes and pure blacks, tending to sep- arate them, especially in the cities. The occupations in which negroes are employed in the cities and in the North have increased and become more standardized in recent years, and intelligence and light color are both probably more in demand in such occupations than in the more simple and isolated labor of rural life. And the life of the so-called underworld of cities, and the fact that inter-racial marriages are permitted in the North, would probably tend to increase the percentage of mulattoes in these places. So while the causes of the distri- bution of mulattoes cannot be definitely stated, it would seem that those mentioned above may reasonably be included among them. It is further worthy of note, that according to the census figures the relative number of mulattoes is increasing. The years in which they were separately classified were 1850, 1860, 1870, 1890 and 1910. And the percentages of mulattoes in the country at large were as follows in those years: 11.2, 13.2, 12.0, 15.2 and 20.9. But this does not mean, as is often sup- posed, that intermixture of whites and negroes has increas- ingly taken place in accordance with these percentages. The relative increase in the number of persons of mixed blood COMPARISON OF SUB-CLASSES OF NEGROES. 87 may have been due to marriage among colored people them- themselves, after a certain degree of racial admixture had oc- curred, without an increase, or even a continuance, of direct crossing between negroes and whites. For the mulattoes and the pure negroes marry each other, and this reduces the original percentage of pure negroes among the total offspring, and increases the percentage of persons of mixed blood. Of course, in the long run, unless there were a new infusion from white stock, the amount of white blood would become very small in any individual. But within the years covered by the census reports there would be enough traces of the white left to cause the individuals possessing them to be classified as having some white characteristics, and the basis of the census classification is the possession of any perceptible trace of white blood. This consideration renders the increase of racial intermixture as indicated by the census figures undoubtedly too high, even if the accuracy of the figures be granted. Views of Other Writers The presence of the mulatto element among the colored population has been taken into account by a number of writers who have dealt with the negro, although it has been neglected by a number of others. In Chapter I was quoted a statement from Strong ('13) to the effect that she divided the negroes tested by her into three classes on the basis of the color of their skin, and found that those of lightest color varied most from the normal, both above and below. Mayo ('13) was aware that the presence of mulattoes might have influenced the school records of the negroes in his study. Pyle ('15) suggested that the superior negroes in his groups might have had a greater proportion of white blood. And Bean ('06) recognized that the brains which he investigated were partly those of mulattoes. Statements from other writers may be given to indicate the attitude of those who have considered the matter from various standpoints. Boas, in concluding a discussion of the American negro problem, writes : "It appears .... that the most important practical questions relating to the negro problem have refer- ence to the mulattoes and other mixed bloods — to their physi- cal tjrpes, their mental and moral qualities, and their vitality. When the bulky literature of this subject is carefully sifted, little remains that will endure serious criticism; and I do 88 THE PSYCHOLOGY OF THE NEGRO. not believe that I claim too much when I say that the whole work on this subject remains to be done. The development of modern methods of research makes it certain that by care- ful inquiry definite answers to our problems may be found. Is it not, then, our plain duty to inform ourselves, that, so far as that can be done, deliberate consideration of observations may take the place of heated discussion of beliefs in matters that concern not only uorselves, but also the welfare of millions of negroes?" (11, pp. 277-278). Stanley Hall says: "The chief event in the history of the Southern negro in the new world is the infiltration of white blood. But for this the negro in mind and body would be so distinct from us that all our problems connected with the race would be vastly simplified. Just how far he has lost his rare racial homogeneity here it is impossible to tell. The extreme minimal estimate that I have found is that one-tenth have some white blood, and one maximal estimate is that two- thirds are partly white. Page thinks that from one-ninth to one-sixth are mixed. DuBois says that two million negroes have some white blood. Most estimates range somewhere between one-fifth and one-half Moreover, the grade of pigmentation is not a sure index of the degree of miscege- nation, and in the veins of some thought purely African prob- ably flows at least a little of the best white blood of the land. .... Thus all the vast psychophysic differences between the two races are bridged, and they possibly fuse with each other by all imperceptible gradations At any rate, men like Fred Douglas, Bishop Payne, Booker Washington, Du Bois Chestnut, Tanner, Dunbar, Thomas and a score of others, are not typical negroes." ('05, pp. 360-361). In a discussion of the general status of the mulatto as compared with the pure negro, Jordan (*13) contends that the former is considerably superior to the latter. The pure negro is claimed to be capable of only the rudiments of civiliza- tion. His powers of attention and reflection are poor. He can draw general conclusions from particular cases only with diflficulty, he is imitative rather than creative, he lacks fore- sight, he has small power of profiting by observation, his character is mobile, and he is guided largely by the instinct of the moment. "The negro cannot undergo mental develop- ment beyond a certain definite maximum. The curious thing is that no attempt is made to establish this opinion on a COMPARISON OF SUB-CLASSES OF NEGROES. 89 scientific basis and to definitely determine the limit of mental development beyond which the law of diminishing returns dictates cessation of effort ; and furthermore that in flat con- tradiction to this common opinion education is planned in apparent utter disregard of it." ('13, pp. 578-579). But with the mulatto the case is different. "Where negro, mulatto / and white are jointly concerned the teachers are unequivocal in their opinion that mental alertness and the development of the higher psychical activities corresponds in degree quite uniformly with the amount of 'white' blood as judged by color of the skin." ('13, p. 577). It is true, argues Jordan, that many mulattoes are inferior, but this is because they come from inferior parents. The way to uplift the negro race is by the proper selection and breeding of mulattoes. A view very different from this is held by Le Bon. He writes: "Doubtless very different races, the black and the white for example, may fuse, but the half-breeds that result constitute a population very inferior to those of which it is sprung, and utterly incapable of creating or even of continu- ing, a civilization. The influence of contrary heredities saps their morality and character. When half-breeds, the off- spring of white men and negroes, have chanced to inherit a superior civilization, as in Saint Domingo, this civilization has speedily been overtaken by the most lamentable degenera- tion. Cross-breeding may be a source of improvement when it occurs between superior and sufficiently allied races, such as the English and the Germans in America, but it always constitutes an element of degeneration when the races, even though superior, are too different." ('98, pp. 52-53). Le Bon quotes Agassiz to the effect that Brazil is undergoing degen- eration on account of the large number of half-breeds in the population, cross-breeding being fatal to the best qualities of whites, blacks and Indians, the peoples concerned. It should be said that this view of Le Bon's is held by only a small minority of those who have discussed the question. Baker's opinion ('08) concerning the relative capacity of the mulatto and the pure negro does not seem to be very decided one way or the other. He points out that a number of leading negroes, as Washington, Du Bois, Douglass, Chest- nutt, Braithwaite, Tanner and Terrell are mulattoes. In- deed, "Most of the leading men of the race to-day in every line of activity are mulattoes." ('08, p. 173). But on the 90 THE PSYCHOLOGY OF THE NEGRO. other hand, such negro leaders as Vernon, Miller, Dunbar, Price and Mason are probably pure-blooded. He claims that to be a mulatto is to be neither a negro nor a white, but is to have the ambitions and hopes of both races. Mulattoes and negroes keep together on account of social pressure from the whites. As to the number of mulattoes, Baker writes: "From my own observation and from talking and correspond- ing with many men who have had superior opportunities for investigation, I think it safe to say that between one-fourth and one-third of the Negroes in this country at the present time have a visible mixture of white blood It is probable that 3,000,000 persons out of the 10,000,000 population are visibly mulattoes." ('08, p. 153). The proportion of mulattoes is stated to be much larger in the North than in the South, and the census figures are mentioned as being unreliable. In a discussion of the political status of the negro, Stone (*08) argues that a distinction should be drawn between the pure negro and the mulatto. "There can no longer be a ques- tion as to the superior intelligence of the mulatto over the Negro — of his higher average potential capacity." ('08, p. 401). The leaders of the colored race, says Stone, have been mulattoes, and this has been true in such places as Jamaica, Santo Domingo, Haiti, South Africa and Liberia, as well as in the United States. The exceptional blacks of pure blood who have attained prominence were generally not descended from true negroes at all. Their ancestry is to be found in those stocks, such as the Fulah and others of high type, that were brought from Africa as slaves in small numbers along with the mass of true negroes. From these abstracts it is evident that there is very little definitely known as to the relative merits of pure negroes and mulattoes. Of opinion, based more or less closely upon ob- servation, there is a great deal. But there has been no seri- ous attempt made to attack the problem from an experi- mental or scientific standpoint. The views of men are uncertain as to the relative abilities of whites and negroes; it is to be expected that they will be much more uncertain when they deal with sub-classes within the colored group. The Classification In the present investigation the negroes tested were divided into four classes on the basis of racial purity as indicated COMPARISON OF SUB-CLASSES OF NEGROES. 91 by color of skin, hair texture and general facial and cranial conformation, the main emphasis being placed upon color. The four classes were pure negroes, negroes three-fourths pure, mulattoes proper, and quadroons. Of course there were probably included in the classification some negroes who did not belong, strictly speaking, in any of these classes. A few octoroons, for example, may have been present, or a few persons who were of seven-eighths pure negro stock. Such cases were placed in the class which seemed to fit their status best, and it is not thought that any constant error resulted from this procedure. It is also probable that errors were made in the classifica- tion. Certain individuals may have been placed in a class too near the pure negroes or too near the pure whites. But here again the incorrect classifications would tend to balance each other, and it is improbable that a constant error in any direc- tion resulted. The effect of the overlapping of the classes would be to lessen the differences found between them.* Con- sequently it follows that in so far as the classification was in- accurate, the real differences were greater than those indicated by the scores. In setting forth the results of the tests, the four classes mentioned are adhered to. But in order to make the classifi- cation still less subject to error, and to secure a larger number of subjects in each class, the pure and the three-fourths pure negroes are grouped together, and the mulattoes and the quad- roons are grouped together, and the results are given sep- arately in all cases for the two resulting classes as well as for the original four. The difference in amount of white blood between these two classes is greater than that between any two classes of the four-class division, and the difference between the scores of the combined pure and three-fourths negroes and the combined mulattoes and quadroons should be correspondingly greater than that between the classes of the original division. This we shall find to be the case. And it *It seems to the writer that probably the greatest chance for erroneous classification occurred between the pure negroes and the ne- groes three-fourths pure. It is not unlikely that differences in color be- tween various stocks of pure negroes may have caused some com- paratively light-colored individuals of this class to be counted as three- fourths pure; and some three-fourths pure individuals of unusually dark native stock may have been classed as pure-blooded. This suppo- sition is borne out by the fact that the differences revealed by the tests between the pure and the three-fourths pure negroes were less than the differences between the other classes. 92 THE PSYCHOLOGY OF THE NEGRO. should be noted in this connection that the class composed of mulattoes and quadroons is very predominantly mulatto, since the quadroons were few in number. This fact obviously makes the difference found between mulattoes and quadroons combined and pure and three-fourths negroes combined less than it really is. It should be said that the classification of the negroes was made by the writer, who had had considerable experience with negroes, at the time the tests were administered. None of the subjects was previously known to him, and he is aware of no circumstance that could have influenced the classifica- tion in addition to the basis previously adopted. The sub- jects seemed to fit into the various classes both naturally and easily, and doubt as to the correctness of a classification was so infrequent as to be negligible. Numbers and Ages The negroes tested in Fredericksburg and Newport News were so few that the results from those cities were not used in this intra-racial study. There was not a sufficient num- ber of each of the four classes in either city to enable results to be computed for any year or any grade. As in the gen- eral comparison between whites and negroes, results from only one subject were not considered. The 319 negroes tested in Richmond fell into the different classes numerically as shown in Tables 33 and 34, which give the numbers by age and sex and by grades. The three high- est and the two lowest years are not used in the comparisons on account of the small number of pupils in them. It may be noted that the largest class is composed of mulattoes, the smallest of quadroons, and that the classes of pure negroes and three-fourths pure negroes are of about equal size. The mulattoes and quadroons together constitute 52 per cent, of the total; the pure and three-fourths negroes together con- stitute 48 per cent, of the total. In the elementary school the mulattoes and quadroons are 46 per cent, of the total number; in the high school the mulattoes and quadroons are 59 per cent, of the total number. There is thus a larger pro- portion of light-colored negroes in the high school than in the elementary school. And in the grades tested as a whole there is a larger percentage of negroes of mixed blood than there is in the city at large, as shown by the census figures. COMPARISON OF SUB-CLASSES OF NEGROES. 93 These facts as to the relative numbers of pure and mixed bloods in the school system indicate that the schools select colored persons of partly white lineage to a greater extent than they select pure negroes, and that as the grades advance the selection becomes more pronounced. This would imply that mulattoes in general are of greater ability and ambition than are pure negroes. And the fact that there is a larger proportion of light-colored negroes in the high school than in the elementary school would partly explain the relatively greater ability of high school negroes that was found in the general comparison between white and colored subjects. Mixed bloods, as will be shown, are of greater ability, and there are proportionately more of them in the upper grades. TABLE XXXIII. Number of Colored Subjects — Richmond — Classified by Racial Purity, Age and Sex Ages 10 11 12 13 14 15 16 17 18 19 21 ' Totals Boys Pure 1 4 8 7 4 4 4 2 2 , , 36 Three-Fourths i 3 2 3 5 4 2 3 1 1 , , 25 Mulattoes 1 8 7 4 7 8 2 3 1 , , 41 Quadroons , , 1 , , 1 2 1 5 Girls Pure 1 4 5 10 8 2 7 1 . . , , 38 Three-Fourths 1 1 10 9 14 6 7 2 4 1 , . 55 Mulattoes 2 4 8 16 12 11 19 10 6 1 1 90 Quadroons , , 8 . , 4 6 2 5 4 4 1 29 Boys Pure and Three-Fourths 1 4 6 11 12 8 6 7 3 3 . . 61 Mulattoes and Quadroons , , 2 8 8 6 8 8 2 3 1 46 Girls Pure and Three-Fourths 1 2 14 14 24 14 9 9 5 1 93 Mulattoes and Quadroons 2 7 8 20 18 13 24 14 10 2 1 119 TABLE XXXIV. Number op Colored Subjects — Richmond — Classified by Racial Purity and Grades Grades 5A 5B 6A 6B 7A 7B lA 2A 2B 3A 4A Totals Pure 10 10 11 5 5 4 12 5 2 3 7 74 Three-Fourths 10 13 10 5 7 5 9 9 5 4 3 80 Mulattoes 20 13 11 9 7 9 14 13 11 12 12 131 Quadroons 2 2 3 3 1 7 4 3 6 3 34 Pure and Three-Fourths 20 23 21 10 12 9 21 14 7 7 10 154 Mulattoes and Quadroons 22 15 14 12 7 10 21 17 14 18 15 165 94 THE PSYCHOLOGY OF THE NEGRO. The ages of the different classes of negroes in the various grades are given in Table 35, and the differences between the age of each class and the age of white pupils of the same grade — the white pupils in question being those tested in Richmond — are given in Table 36. This latter table shows that the pure negroes were .69 of a year older than the white pupils; the three-fourths pure negroes were .32 of a year older; the mulattoes were .29 of a year older; and the quad- roons were .07 of a year younger. The pure and three-fourths negroes combined were .51 of a year older than the whites; the mulattoes and quadroons combined were .23 of a year older. These figures are significant. They show that the darker negroes were slower in reaching a given grade than were those of lighter color. The differences are small, and are subject to variation from grade to grade. But the aver- age differences, which are here quoted, are uniformly greater for the darker negroes, and the probable errors are not large enough to invalidate the significance of this uniformity. TABLE XXXV. Ages of the Colored Subjects — Richmond — Classified by Racial Purity and Grades Grades 5A 5B 6A 6B 7A 7B lA 2A 2B 3A 4A Pure Av. 12.7 13.2 12.8 14.2 14.8 14.7 15.4 17.0 16.5 16.0 17.8 A.D. .9 .6 .6 .3 .3 .7 .8 .4 .5 1.3 .7 Three-Fourths Av. 11.7 13.1 13.0 13.4 13.7 14.4 14.3 16.6 16.6 17.7 17.0 A.D. 1.0 1.0 .8 1.5 .6 .6 .6 .6 1.1 .9 1.3 Mulattoes Av. 12.6 12.4 12.8 13.2 13.8 14.6 15.2 15.8 16.5 16.6 17.6 A.D. 1.1 .9 .6 .8 1.3 .8 .5 .8 1.0 .5 1.2 Quadroons Av. 12.0 12.0 11.6 14.0 . . . . 14.0 16.0 16.6 17.0 17.6 A.D. 1.0 1.0 .9 3 1.0 .9 .7 .9 Pure and Three-Fourths Av. 12.2 13.1 12.9 13.8 14.1 14.5 14.9 16.7 16.5 17.0 17.6 A.D. 1.0 .8 .7 .9 .7 .7 .8 .6 .9 1.1 .9 Mulattoes and Quadroons Av. 12.5 12.4 12.5 13.4 13.8 14.5 14.8 15.8 16.5 16.7 17.6 A.D. 1.1 .9 .9 .8 1.3 .9 .7 .9 1.0 .6 1.2 TABLE XXXVI. Ages — Difference in Years Between the White and Each of the Four Classes of Colored Subjects Tested — Richmond (Minus signs indicate that the colored subjects are of greater age.) Grades 5A 5B 6A 6B 7A 7B lA 2A 3A 4A Av. P.E. Pure Negroes ^ „ «« ^c —.3 —.7 .6 —1.0 —1.5 —1.1 —1.0 —1.6 .5 — .8 —.69 .15 Three-Fourths .7 _.6 .4 _ .2 — .4 .8 .1 —1.2 —1.2 —.32 .14 Mulattoes _.2 .1 .6 — .5 —1.0 _ .8 — .4 — .1 — .6 —.29 .10 COMPARISON OF SUB-CLASSES OF NEGROES. 95 5A 5B 6A 6B 7A 7B lA 2A 3A 4A Av. P.E. Quadroons .4 .5 1.8 — .8 4 — .6 — .5 — .6 .07 J21 Pure and Three-Fourths M 1 ./^ ~-^A r. •^- .6 - .8 - .9 - .5 -1.3 - .5 - .6 -.51 .09 Mulattoes and Quadroons —.1 .1 .9 — 2 — .5 — .9 — .4 — .4 — .2 — .6 —.23 .09 Standing in the Tests The tests used in this comparison are the mixed relations, Tests I and II, and the completion. These tests revealed marked differences between whites and negroes as a whole, and they are therefore well adapted to bring out any differ- ences that may exist between the various classes of negroes. The maze and cancellation tests did not show any considerable differences in the general inter-racial comparison, and they are consequently not employed here. ^ In setting forth the results of the tests the same order is followed as in the preceding chapter. In the case of each test, first are given the actual scores, with their average devia- tions, classified by age and sex and by grades. Then appear graphs based upon these scores. The graphs are drawn only for the two-division classification — that of pure and three- fourths negroes combined and mulattoes and quadroons com- bined. The scores of the white pupils tested in Richmond are included in the graphs for the sake of comparison. Fol- lowing the graphs are the tables in which the comparisons are made. In each instance the score of the colored subjects is compared with that of the corresponding white pupils. The actual differences between the scores are given; the percentage of the score of the whites that was obtained by the negroes next appears; and lastly are shown the per- centages of the two-division classification that reach or ex- ceed the average score of the whites. The averages of the various ages and grades, with their probable errors, appear in each table of comparison. Tables 37 and 38 give the scores in the mixed relations test. Figures 16-21 present these scores in graphic form. Tables 39-43 make the comparisons. From these latter tables the following facts appear. 96 THE PSYCHOLOGY OF THE NEGRO. TABLE XXXVII. Mixed Relations Test — Scores of Colored Subjects — Richmond — Classified by Racial Purity, Age and Sex Ages 10 11 12 13 14 15 16 17 18 19 Test I. Boys — Pure Av 6.5 12.3 9.7 10.5 19.3 18.7 20.0 16.5 A.D 1.0 4.1 3.2 1.5 8.3 5.2 4.0 10.5 Boys — Three-Fourths Av. .. 9.0 9.0 17.0 18.5 6.5 21.0 19.3 A.D. .. 2.6 1.0 6.0 5.5 2.5 0. 10.3 Boys — Mulattoes Av 12.5 7.4 12.5 17.2 21.0 32.5 19.6 A.D 6.0 3.5 4.5 8.5 9.5 2.5 12.6 Boys — Quadroons Av 28.0 A.D Girls — Pure Av 9.6 10.5 12.6 15.1 17.0 21.4 A.D 4.0 6.0 4.4 7.5 5.0 6.1 Girls — Three-Fourths Av 9.8 10.6 11.4 20.8 23.4 16.0 16.5 A.D 4.6 3.7 6.1 7.8 6.7 4.0 6.0 Girls — Mulattoes Av. 9.5 12.0 12.5 14.6 15.7 18.1 20.1 29.5 14.8 A.D. 7.5 3.0 6.1 6.9 9.9 9.4 7.2 4.7 6.5 Girls — Quadroons Av. .. 9.6 .. 10.7 23.0 20.0 23.4 30.7 25.2 A.D. .. 5.3 .. 7.2 8.3 12.0 6.4 4.2 9.7 Test II. Boys — Pure Av 7.7 13.0 12.8 18.0 24.0 26.5 36.5 28.5 A.D 2.7 3.5 8.8 7.0 7.0 10.5 1.5 .5 Boys — Three-Fourths , Av. .. 11.3 9.0 17.3 18.8 4,5 22.0 33.3 A.D. .. 7.0 0. 7.0 10.8 1.5 7.0 3.0 Boys — Mulattoes Av 14.0 9.8 15.0 27.5 24.5 39.5 27.6 A.D 8.2 3.4 8.0 7.1 10.0 .5 8.0 Boys — Quadroons . Av 25.5 A.D 13.5 Girls — Pure . Av 9.0 8.2 13.7 19.8 28.5 26.8 A.D 3.3 2.2 7.5 8.8 7.5 9.7 Girls — Three-Fourths Av 9.8 15.2 11.2 26.3 29.7 22.0 20.2 A.D 4.8 6.6 7.5 7.3 6.7 4.0 5.2 Girls — Mulattoes Av. 13.0 14.0 12.1 11.6 20.5 20.6 26.4 35.3 19.8 A.D. 5.0 5.0 5.3 7.6 12.2 10.0 6.6 3.7 10.8 Girls — Quadroons Av. .. 15.5 .. 8.6 24.6 27.0 30.8 37.5 30.7 A.D. .. 3.5 .. 4.6 12.6 10,0 5.2 1.0 8.2 Test I. Boys — Pure and Three-Fourths Av. .. 10.5 7.3 13.6 12.9 7.8 19.7 19.0 20.3 16.5 A.D. .. 3.5 1.3 4.2 6.0 2.8 6.7 7.4 3.0 10.5 Boys — Mulattoes and Quadroons Av. .. 13.5 12.5 9.5 15.6 17.2 21.0 32.5 19.6 A.D. .. 4.5 6.0 4.7 7.4 8.5 9.5 2.5 12,6 Girls — Pure and Three-Fourths 13 10.6 4.4 14 11.9 5.3 15 18.0 8.3 16 22.0 6.2 17 20.2 5.5 18 17.2 5.4 19 13.7 7.2 18.1 9.9 18.4 9.8 20.8 6.9 29.8 4.7 19.0 10.0 29.5 .5 14.1 4.7 15.3 10.0 11.2 7.5 23.3 7.0 29.8 8.2 36.6 1.3 32.0 4.6 'l3.2 7.2 18.5 10.8 28.4 7.0 24.5 10.0 39.5 .5 27.6 8.0 •• 12.7 5.8 12.2 7.7 22.6 8.0 29.4 6.8 25.7 8.5 19.8 4.8 •• 11.1 5.0 21.9 4.5 21.6 5.3 27.3 7.3 ] 35.9 LO.O ] 24.2 LO.l 28.0 6.6 COMPARISON OF SUB-CLASSES OF NEGROES. 97 Ages 10 11 12 Av. .. 17.5 9.7 A.D. .. 3.5 4.3 Girls — Mulattoes and Quadroons Av. 9.5 11.0 12.5 A.D. 7.5 4.2 6.1 Test II. Boys — Pure and Three-Fourths Av. .. 12.0 8.0 A.D. .. 6.0 2.8 Boys — Mulattoes and Quadroons Av. .. 14.0 14.0 A.D. .. 12.0 8.2 Girls — Pure and Three-Fourths Av. .. 25.5 9.6 A.D. .. .5 4.5 Girls — Mulattoes and Quadroons Av. 13.0 14.5 12.1 3.3 11.0 0. A.D. TABLE XXXVIII. Mixed Relations Test — Scores of Colored Subjects — Richmond — Classified by Racial Purity and Grades Grades 5A 5B 6A 6B 7A 7B lA 2A 2B 3A 4A Test I. Pure Av. 7.6 10.5 12.4 12.2 13.3 11.7 18.4 17.4 21.0 28.6 17.5 A.D. 2.6 3.8 4.2 3.7 4.3 5.2 7.7 4.4 1.0 3.3 6.5 Three-Fourths Av. 9.3 8.6 13.6 17.0 8.7 13.4 17.7 20.0 22.0 22.0 22.3 A.D. 3.5 1.6 5.6 4.0 5.5 6.4 4.1 7.1 7.6 6.6 10.3 Mulattoes Av. 9.8 12.3 13.1 12.8 9.0 11.6 23.2 22.0 16.4 30.0 24.0 A.D. 3.9 4.7 7.6 6.4 4.0 5.5 6.5 9.8 6.0 3.9 10.3 Quadroons Av. 4.5 9.0 12.3 13.0 .. .. 27.3 22.2 13.3 32.1 29.6 A.D. 2.5 1.0 7.0 2.0 .. .. 4.6 8.2 5.0 3.1 4.6 Test II. Pure Av. 8.7 9.8 12.5 15.2 18.2 11.5 22.5 28.4 26.0 35.0 28.4 A.D. 4.5 4.8 4.5 4.2 7.8 5.0 10.0 4.4 10.0 1.3 7.5 Three-Fourths Av. 12.6 8.9 12.1 16.0 14.4 10.6 24.6 27.2 29.0 34.2 24.3 A.D. 6.8 4.5 6.1 8,8 7.5 6.6 8.4 7.5 6.4 4.7 9.6 Mulattoes ^ „„ ^ Av. 9.1 12.1 15.7 14.0 17.4 13.4 28.6 28.3 21.8 35.6 30.5 A.D. 4.2 7.6 8.1 8.2 7.2 7.1 6.6 7.3 7.1 2.8 8.6 Quadroons „ ^^ ^ Av. 4.0 9.0 22.5 8.0 .. .. 35.1 29.7 23.0 36.1 34.6 A.D. 0. 3.0 3.5 2.6 .. .. 2.4 6.7 8.0 1.5 4.6 Test I. Pure and Three-Fourths ^ ^ ^ „ Av. 8.4 9.4 13.0 14.2 10.1 12.6 18.1 19.0 21.7 25.3 19.0 A.D. 3.0 2.7 4.9 4.0 5.9 6.0 5.8 5.9 5.5 5.6 8.0 Mulattoes and Quadroons^ H.g 13.0 12.9 9.0 13.1 24.5 22.0 15.7 30.7 25.1 A.D. 3.9 4.0 7.4 5.6 4.0 6.9 6.3 9.4 5.7 3.6 9.4 Test II. Pure and Three-Fourths 98 THE PSYCHOLOGY OF THE NEGRO. Grades 5A 5B 6A 6B 7A 7B lA 2A 2B 3A 4A Av. 10.5 9.3 12.3 15.6 16.0 11.0 23.4 27.6 28.1 34.5 27.2 A.D. 5.5 4.8 5.2 6.4 8.0 6.2 9.3 6.6 7.5 3.4 8.4 Mulattoes and Quadroons Av. 8.9 11.7 16.7 12.5 17.4 13.7 30.8 28.6 22.0 35.8 31.3 A.D. 4.1 6.8 8.0 7.0 7.2 6.7 5.7 7.3 7.3 2.5 7.9 In Test I, in the case of boys, classified by age, the pure ne- groes obtained 58.8 per cent, of the score of the whites; the three-fourths pure negroes obtained 72.2 per cent, of the score of the whites ; and the mulattoes obtained 78.5 per cent, of the white score.* When the boys are grouped into two classes, pure and three-fourths combined and mulattoes and quadroons combined, the former class is shown to have ob- tained 62.0 per cent, of the score of the whites while the latter class obtained 83.5 per cent, of the white score. The girls in Test I, classified by age, obtained the following percentages of the white score: Pure negroes, 68.0; three- fourths pure negroes, 71.3; mulattoes, 87.5; quadroons, 99.0. When the girls are arranged in two classes the results are as follows; The pure and three-fourths negroes combined scored 72.0 per cent, as high as the whites; the mulattoes and quadroons combined scored 89.3 per cent, as high as the whites. The grade comparison for Test I shows that the pure ne- groes, boys and girls together, scored 73.3 per cent, as high as the whites ; the three-fourths pure negroes scored 74.6 per cent, as high; the mulattoes scored 81.6 per cent, as high; the quadroons scored 87.9 per cent, as high as the whites. The two-division classification by grades shows that the pure and three-fourths negroes obtained 72.9 per cent, of the white score, and that the mulattoes and quadroons obtained 82.8 per cent, of the white score. The percentages of negroes reaching or exceeding the aver- age of the whites were as follows in Test I : Boys — ^Pure and three-fourths combined, 17.0; Mulattoes and quadroons com- bined, 38.5. Girls — ^Pure and three-fourths combined, 25.5; *There was not a sufficiently large number of quadroon boys to be included as a separate class in the tables of comparison. There were only five such boys in all, and only two of them were in any one age group, namely, age 14. These two scored higher than any of the other classes of negroes, but of course a result from such a small number is practically worthless. The quadroon boys are included in the groups of combined mulattoes and quadroons, however, and in the grade groups of quadroons, which contain both boys and girls. COMPARISON OF SUB-CLASSES OF NEGROES. 99 Mulattoes and quadroons combined, 41.3. Grades — ^Pure and three-fourths combined, 25.4; Mulattoes and quadroons com- bined, 38.4. Fig. 16. Mixed Relations Test I — Scores of Whites, Mulattoes and Quadroons Combined, and Pure and Three-Fourths Negroes Combined — Boys — Richmond.* *The white, the shaded and the black columns indicate the scores of the whites, the mulattoes and quadroons combined, and the pure and three-fourths negroes combined, respectively. A&E Fig. 17. Mixed Relations Test I — Scores of Whites, Mulattoes and Quadroons Combined, and Pure and Three-Fourths Negroes Combined — Girls — Richmond Fig. 18. Mixed Relations Test I — Scores of Whites, Mulattoes and Quadroons Combined, and Pure and Three-Fourths Negroes Combined — Grades — Richmond. 100 THE PSYCHOLOGY OF THE NEGRO. AfrE Fig. 19. Mixed Relations Test II — Scores of Whites, Mulattoes and Quadroons Combined, and Pure and Three-Fourths Negroes Combined — Boys — Richmond. Fig. 20. Mixed Relations Test II — Scores of Whites, Mulattoes and Quadroons Combined, and Pure and Three-Fourths Negroes Combined — Girls — Richmond. Fig. 21. Mixed Relations Test II — Scores of Whites, Mulattoes and Quadroons Combined, and Pure and Three-Fourths Negroes Combined — Grades — Richmond. These figures show that in Test I the size of the score made by the negroes varies directly with the amount of their white blood. As judged by the averages, this is true without excep- COMPARISON OF SUB-CLASSES OF NEGROES. 101 tion for both boys and girls and in both the age and the grade classifications. The size of the probable errors and the con- stancy of the results renders the differences reliable. In the two-division classification the differences are larger than in the four-division classification, as was to be expected, and the results here are especially reliable when the size of the proba- ble errors is considered. TABLE XXXIX. Mixed Relations Test — Difference Between Scores of the White AND Each of the Four Classes of Colored Subjects, by Age and Sex — Richmond (Minus signs indicate greater scores by the colored subjects.) Ages 12 13 14 15 16 17 Av. P.E. Test I. Boys— Pure 10.0 6.0 8.3 9.6 8.5 6.9 8.2 .4 Boys— Three-Fourths 7.5 1.3 —.5 13.6 6.8 6.3 5.8. 1.3 Boys— Mulattoes 4.0 10.9 5.5 2.9 6.8 —6.9 3.9 1.4 Girls— Pure 9.2 7.5 10.1 7.9 8.0 —1.6 6.8 1.0 Girls— Three-Fourths 9.0 7.4 11.3 2.2 1.6 3.8 5.9 1.2 Girls— Mulattoes 6.3 3.4 7.0 4.9 4.9 —9.7 2.8 1.5 Girls— Quadroons . . 7.3 —.3 3.0 1.6 —10.9 .1 1.8 Test II. Boys— Pure 11.8 8.3 11.3 4.0 9.0 8.1 8.7 .7 Boys~Three-Fourthsl0.5 4.0 5.3 17.5 11.0 1.3 8.3 1.6 Boys— Mulattoes 5.5 11.5 9.1 —5.5 8.5 —4.9 4.0 2.2 Girls— Pure 13.4 13.3 11.0 8.8 3.2 1.0 8.4 1.5 Girls-Three-Fourths 12.6 6.3 13.5 2.3 2.0 5.8 7.1 1.4 Girls— Mulattoes 10.3 9.9 4.2 8.0 5.3 —7.5 5.0 1.6 Girls— Quadroons .. 12.9 .1 1.6 .9 —9.7 1.2 1.8 Test I. Boys Pure and Three-Fourths 9.2 4.7 5.1 12.3 8.1 6.6 7.7 .8 Mulattoes and Quadroons 4.0 8.8 2.4 2.9 6.8 —6.9 3.0 1.2 Girls Pure and Three-Fourths 9.? 7.4 10.8 5.0 3.0 —.4 5.8 1.2 Mulattoes and ^ . Quadroons 6.3 4.3 4.6 4.6 4.2 —10.0 2.3 1.4 Test II. Boys Pure and ^ _ ^ „ „ „ _ Three-Fourths 11.5 7.2 8.8 10.8 9.7 4.8 8.8 .7 Mulattoes and ^ , ^ ^ „ „ _ . Quadroons 5.5 8.1 5.6 —6.4 8.5 —4.9 2.7 2.0 Girls PurG &Tld Three-Fourths 12.8 8.8 12.5 6.0 2.3 2.1 7.4 1.4 Mulattoes and ^ ^ ^„ ^ ^ „, ._ ,_ Quadroons 10.3 10.4 2.8 7.0 4.4 —8.1 4.5 1.7 102 THE PSYCHOLOGY OF THE NEGRO. TABLE XL. Mixed Relations Test — Difference Between Scores op thp White AND Each of the Four Classes of Colored Subjects, by Grades — Richmond (Minus signs indicate greater scores by the colored subjects.) Grades 5A 5B 6A 6B 7A 7B lA 2A 3A 4A Av. Test I. Pure 5.5 3.4 4.4 4.6 5.1 13.2 7.1 10.6 —5.5 5.8 5.4 Three-Fourths 3.8 5.3 3.2 — .2 9.7 11.5 7.8 8.0 1.1 1.0 5.1 Mulattoes 3.3 1.6 3.7 4.0 9.4 13.3 2.3 6.0 —6.9 — .7 3.6 Quadroons 8.6 4.9 4.5 3.8 . . , . —1.8 5.8 —9.0 —6.3 1.3 Test IL Pure 6.1 8.2 7.3 7.4 3.6 17.9 6.5 5.2 —3.2 1.8 6.1 Three-Fourths 2.2 9.1 7.7 6.6 7.4 18.8 4.4 6.4 —2.4 5.9 6.6 Mulattoes 5.7 5.9 4.1 8.6 4.4 16.0 .4 5.3 —3.8 — .3 4.6 Quadroons 10.8 9.0 —2.7 14.6 , , , , —6.1 3.9 —4.3 —4.4 2.6 Test I. Pure and Three-Fourths 4.7 4.5 3.8 2.6 8.3 12.3 7:4 9.0 —2.2 4.3 5.5 Mulattoes and Quadroons 3.8 2.1 3.8 3.9 9.4 11.8 1.0 6.0 —7.6 —1.8 3:2 Test n. Pure and Three-Fourths 4.3 8.7 7.5 7.0 5.8 18.4 5.6 6.0 —2.7 3.0 6.4 Mulattoes and Quadroons 5.9 6.3 3.1 10.1 4.4 15.7 —1.8 5.0 —4.0 —1.1 4.4 TABLE XLI. Mixed Relations Test — Percentage of the Scora; op the White Obtained by Each op the Four Classes of Colored Subjects, BY Age and Sex — Richmond Ages 12 13 14 15 16 17 Av. P.E. Test I. Boys — Pure 37 67 54 52 70 73 58.8 3.8 Boys — Three-Fourths 53 93 103 32 76 76 72.2 7.0 Boys— -Mulattoes 75 39 69 85 76 127 78.5 6.3 Girls — Pure 52 58 56 66 68 108 68.0 4.5 Girls- Three-Fourths 53 59 51 90 94 81 71.3 5.9 Girls — Mulattoes 67 81 70 79 80 148 87.5 7.0 Girls — Quadroons Test n. Boys — Pure •• 59 101 87 94 154 99.0 8.7 38 60 53 82 73 77 63.8 4.5 Boys — Three-Fourths 45 81 78 20 67 96 64.5 7.3 Boys — Mulattoes 71 45 62 125 74 114 81.8 8.7 Girls — Pure 39 37 56 70 90 96 64.7 7.3 Girls — Three-Fourths 43 70 46 92 94 79 70.7 6.3 Girls — Mulattoes 53 53 83 72 83 127 78.5 6.6 Girls — Quadroons , , 39 100 94 97 135 93.0 8.4 Test I. Boys — Pure and Three-Fourths 42 74 72 38 71 75 62.0 5.2 Mulattoes and Quadroons 75 51 87 85 76 127 83.5 5.6 Girls — Pure and Three-Fourths 52 59 53 78 88 102 72.0 5.9 Mulattoes and Quadroons 67 76 80 80 83 150 89.3 7.0 COMPARISON OF SUB-CLASSES OF NEGROES. 103 Ages 12 13 14 15 16 17 Av. P.E. Test II. Boys — Pure and Three-Fourths 39 66 63 51 71 86 62.7 4.2 Mulattoes and Quadroons Girls — Pure and 71 61 77 129 74 114 87.7 8.0 Three-Fourths 42 58 50 79 93 92 69.0 6.6 Mulattoes and Quadroons 53 50 89 76 86 129 80.5 7.3 TABLE XLII. Mixed Relations Test — Percentage op the Score op the Whith Obtained by Each of the Four Classes op Colored Subjects, by Grades — Richmond Grades Test I. Pure Three-Fourths Mulattoes Quadroons Test II. Pure Three-Fourths Mulattoes Quadroons Test I. Pure and Three-Fourths Mulattoes and Quadroons Test II. Pure and Three-Fourths Mulattoes and Quadroons 5A 5B 6A 6B 7A 7B lA 2A 3A 4A Av. P.E. 58 76 74 73 72 47 72 62 124 75 73.3 2.8 71 62 81 101 46 54 69 71 95 96 74.6 3.9 75 89 78 76 48 47 91 79 130 103 81.6 4.4 34 65 74 78 . . . . 107 79 139 127 87.9 8.1 59 54 63 68 84 38 78 85 110 94 73.3 4.4 85 49 61 71 66 35 85 81 107 80 72.0 4.2 62 67 79 63 80 45 99 84 112 101 79.2 4.2 28 50 113 37 . . . . 121 89 113 115 83.2 10.2 64 68 78 85 54 51 70 68 110 81 72.9 3.1 71 85 78 77 48 53 96 79 133 108 82.8 4.7 71 52 62 70 74 37 81 82 108 90 72.7 3.6 61 65 84 56 80 46 106 85 112 104 79.9 4.7 TABLE XLIIL Mixed Relations Test — Percentage op Each of Two Classes op Colored Subjects Reaching or Exceeding the Average op thb White, by Age and Sex and by Grades — Richmond Ages Test I. Boys — Pure and Three-Fourths Mulattoes and Quadroons Girls — Pure and Three-Fourths Mulattoes and Quadroons Test II. Boys — Pure and Three-Fourths Mulattoes and Quadroons Girls — Pure and Three-Fourths Mulattoes and Quadroons 12 13 14 15 16 17 27 36 25 14 25 12 40 29 25 100 8 23 8 25 33 56 17 28 39 38 33 93 18 25 12 17 43 25 12 33 86 25 100 8 7 12 36 56 44 12 21 56 38 37 93 Av. P.E. 17.0 4.2 38.5 7.3 25.5 4.2 41.3 5.9 19.2 3.4 46.8 10.1 27.2 6.3 42.8 7.3 104 THE PSYCHOLOGY OF THE NEGRO. Grades 5A 5B 6A 6B 7A 7B lA 2A 3A 4A .. .. Test I. Pure and Three-Fourths 15 14 38 29 20 17 14 67 40 25.4 3.6 Mulattoes and Quadroons 27 31 36 18 20 50 41 94 67 38.4 5.2 Test II. Pure and Three-Fourths 26 19 19 20 25 48 21 86 40 30.4 4.2 Mulattoes and Quadroons 10 20 31 17 43 76 29 89 60 37.5 6.0 We may conveniently average the percentages obtained by the boys, girls and grades in each class. When this is done the scores for the four classes of negroes, in terms of per- centages of the score of whites, are as follows : Pure negroes, 66.7.; negroes three-fourths pure, 72.7; mulattoes, 82.5; quad- roons, 93.4. The scores of the two-division classification, in the same terms, are: Pure and three-fourths negroes, 69.0; mulattoes and quadroons, 85.2. Averaging the percentages of boys, girls and grades that reached or exceeded the average score of whites, we find that 22.6 per cent, of the combined pure and three-fourths negroes, and 39.4 per cent, of the com- bined mulattoes and quadroons reached the white average. In Test II the same situation occurs. The percentages of the score of the whites obtained by the pure, the three-fourths pure and the mulatto boys were 63.8, 64.5 and 81.8, respec- tively. For the girls, the percentages obtained by the pure negroes, the negroes three-fourths pure, the mulattoes and the quadroons were, respectively, 64.7, 70.7, 78.5 and 93.0. For the grades, the percentages, in the same order, were 73.3, 72.0, 79.2 and 83.2.* The combined pure and three-fourths pure negroes and the combined mulattoes and quadroons scored, in Test II, the fol- lowing respective percentages of the score of the whites: Boys— 62.7 and 87.7 ; Girls— 69.0 and 80.5 ; Grades— 72.7 and 79.9. *It should be noted that in the comparison by ^ades in Test II the pure negroes scored 1.3 per cent, higher than those three-fourths pure. In the grade comparison of the completion test the pure negroes also scored higher, by 2.0 per cent., than the three-fourths negroes. These are the only instances in which subjects with a greater amount of white blood were inferior to those with a lesser amount. As was previously pointed out, comparisons by grades are not as likely to reveal marked or reliable differences as are comparisons by ages; and as was also pointed out, the classes of pure and three-fourths pure negroes may not have been as well separated as the other classes. COMPARISON OF SUB-CLASSES OF NEGROES. 105 The percentages of negroes reaching or exceeding the aver- age of the whites in Test II were as follows : Boys — Pure and three-fourths, 19.2; Mulattoes and quadroons, 46.8. Girls — Pure and three-fourths, 27.2; Mulattoes and quadroons, 42.8. Grades — Pure and three-fourths, 30.4; Mulattoes and quad- roons, 37.5. If we average for Test II the percentages obtained by the boys, girls and grades in each class, we find that the pure negroes obtained 67.3 per cnt. of the score of the whites; that the three-fourths pure negroes obtained 69.1 per cent, of the score of the whites; that the mulattoes obtained 79.8 per cent, of the white score ; and that the quadroons obtained 88.1 per cent, of the white score. The pure and three-fourths negroes, boys, girls and grades, scored 68.1 per cent, as high as the whites; the mulattoes and quadroons, boys, girls and grades, scored 82.7 per cent, as high as the whites. Aver- aging the percentages of boys, girls and grades that reached or exceeded the average white score, we find that 25.6 per cent, of the pure and three-fourths pure negroes reached the white average; and that the white average was reached by 42.4 per cent, of the mulattoes and quadroons. The results of the completion test appear in Tables 44 and 45 and in Figures 22-24. The comparisons are made in Ta- bles 46-50. These latter tables show the following results: In the case of boys, the pure negroes obtained 65.3 per cent, of the score of the whites; the three-fourths pure ne- groes obtained 76.2 per cent. ; and the mulattoes obtained 79.2 per cent. For girls, the percentages for the four classes of negroes, pure, three-fourths pure, mulattoes and quadroons, were, respectively, 74.3, 77.7, 81.8 and 95.8. For grades, the percentages were: Pure negroes, 81.4; three-fourths pure, 79.4; mulattoes, 82.9; quadroons, 92.2. When the pure and three- fourths pure negroes and the mulattoes and quadroons are grouped together, it appears that the pure and three-fourths pure boys scored 68.7 per cent, as high as the whites, and that the mulatto and quadroon boys scored 82.0 per cent, as high as the whites. The girls, purs and three-fourths pure, obtained 77.3 per cent, of the white score ; and the girls, mu- lattoes and quadroons, obtained 83.5 per cent, of the white score. In the case of the grades, the pure and three-fourths pure negroes scored 80.2 per cent, as high as the whites, and the mulattoes and quadroons scored 83.4 per cent, as high. 106 THE PSYCHOLOGY OF THE NEGRO. TABLE XLIV. Completion Test — Scores op Colored Subjects — Richmond — Classi- fied BY Racial Purity, Age and Sex Ages 10 11 12 13 14 15 16 17 18 19 Boys Pure Av. , . 12.5 12.7 15.5 12.0 19.2 20.0 34.0 31.5 A.D. , , 1.0 3.2 5.7 3.0 2.2 5.0 0. 3.5 Three-Fourths Av. , , 8.6 15.0 17.3 14.7 10.2 16.0 36.3 A.D. 2.6 5.0 3.0 6.7 3.7 6.0 3.6 Mulattoes Av. , , 13.2 11.4 15.5 20.7 23.6 32.0 29.6 , , A.D. , , , , 2.5 5.5 6.5 13.2 11.3 1.0 8.6 Quadroons Av. , . . . 19.5 , , . . . . A.D. , , ^ , , , , , 5.5 , , , , , , , , Girls Pure Av. , , , , 12.7 13.8 16.4 19.2 23.0 27.0 ^ , , ^ A.D. 3.2 4.4 3.5 4.5 2.0 9.4 . , Three-Fourths Av. 13.5 15.4 15.0 22.6 27.1 22.5 23.5 A.D. , , , , 1.3 2.8 4.7 3.3 4.1 2.5 4.5 , ^ Mulattoes Av. 12.0 13.7 15.1 14.3 21.0 21.1 22.7 29.1 22.5 , ^ A.D. 5.0 3.2 4.8 4.5 8.5 5.4 5.6 6.1 2.1 Quadroons Av. , , 18.6 , , 13.7 21.0 35.5 24.6 30.5 27.7 , , A.D. 1.3 ^ , 2.2 7.0 2.5 4.6 1.0 3.2 Poys — Pure and \ Three-Fourths Av. , , 8.7 13.3 14.0 15.2 11.1 18.1 27.0 33.6 30.6 A.D. , ^ 2.2 2.3 3.4 6.0 3.3 3.5 8.8 .6 2.6 Mulattoes and Quadroons Av. 9.0 13.2 13.1 16.8 21.2 23.6 32.0 29.0 , . A.D. , , 1.0 2.5 6.6 6.1 12.2 11.3 1.0 8.6 ^ ^ Girls — Pure and Three-Fourths Av. 20.5 13.2 14.8 15.5 20.7 26.2 26.0 22.6 A.D. , , 3.5 1.8 3.4 4.2 4.1 3.7 8.4 4.2 . . Mulattoes and Quadroons Av. 12.0 15.8 15.1 14.2 21.0 23.3 23.1 29.5 24.6 29.0 A.D. 5.0 3.1 4.8 3.9 8.0 6.5 5.5 4.7 3.6 0. TABLE XLV. Completion Test — Scores op Colored Subjects — Richmond — Classi- fied BY Racial Purity and Grades Grades Pure Av. A.D. Three-Fourths Av. A.D. Mulattoes Av. A.D. Quadroons Av. A.D. Pure and Three-Fourths Av. A.D. Mulattoes and [ Quadroons Av. A.D. 5A 5B 6A 6B 7A 7B lA 2A 2B 3A 4A 10.6 14.0 14.2 19.5 17.8 17.2 17.8 23.8 29.0 29.3 26.8 2.4 1.4 3.7 2.0 5.2 3.7 4.3 4.4 4.0 7.6 9.7 11.9 13.4 13.3 13.5 14.8 16.0 21.5 27.5 25.8 28.7 27.6 5.5 3.0 3.7 2.5 4.2 5.2 5.6 5.8 5.2 3.7 2.0 10.5 12.0 14.9 18.2 16.2 17.2 23.2 28.0 21.0 27.2 31.5 3.6 2.7 4.9 4.0 3.7 4.0 8.7 6.8 4.3 5.7 5.6 16.0 16.5 15.3 11.6 . . . . 26.2 28.5 24.6 30.0 29.3 1.0 2.5 3.6 1.3 .. .. 2.2 7.0 2.6 2.0 .3 11.2 13.6 13.8 16.5 16.0 16.5 19.4 26.2 26.7 29.0 27.1 4.1 2.4 3.6 3.2 4.7 4.3 4.7 5.5 5.0 5.1 7.1 11.0 12.6 15.0 16.5 16.2 16.5 24.2 28.1 21.7 28.1 31.0 3.7 3.0 4.4 4.5 3.7 4.3 6.7 6.8 4.2 4.5 4.8 COMPARISON OF SUB-CLASSES OF NEGROES. 107 The percentages of negroes reaching or exceeding the aver- age score of the whites were as follows: Boys — Pure and three-fourths pure, 19.0; Mulattoes and quadroons, 44.S. Girls — Pure and three-fourths pure. 23.2; Mulattoes and quadroons, 31.5. Grades — Pure and three-fourths pure, 25.9; Mulattoes and quadroons, 27.9. A&E Fig. 22. Completion Test — Scores of Whites, Mulattoes and Quad- roons Combined, and Pure and Three-Fourths Negroes Combined — Boys — Richmond. WNfQP-TlW 12 I 13 23. Completion Test — Scores of Whites, Mulattoes and Quad- roons Combined, and Pure and Three-Fourths Negroes Combined — Girls — Richmond. Fig Fig. 24. Completion Test — Scores of Whites, Mulattoes and Quad- roons Combined, and Pure and Three-Fourths Negroes Combined- Grades — Richmond. 108 THE PSYCHOLOGY OF THE NEGRO. If we average the percentages for boys, girls and grades, we find that the pure negroes scored 73.7 per cent, as high as the whites, the three-fourths pure negroes scored 77.8 per cent, as high, the mulattoes scored 81.3 per cent, as high, and the quadroons scored 94.0 per cent, as high as the whites. The pure and three-fourths pure negroes, combined, obtained 75.4 per cent, of the white score; the mulattoes and quad- roons, combined, obtained 83.0 per cent, of the white score. In terms of the percentage of negroes reaching or exceeding the average of the whites, the results for boys, girls and grades are: Pure and three-fourths pure negroes, 22.7 per cent.; Mulattoes and quadroons, 34.7 per cent. TABLE XLVI. Completion Test — Difference Between Scores of the White and Each of the Four Classes of Colored Subjects, by Age and Sex — Richmond (Minus signs indicate greater scores by the colored subjects.) Ages 12 13 14 15 16 17 Av. P.E. Boys Pure 6.2 6.0 5.7 11.4 11.5 10.7 8.6 .9 Three-Fourths 3.7 1.4 6.5 13.2 14.7 —5.6 5.6 2.0 Mulattoes 5.5 7.3 5.7 2.7 7.1 —1.3 4.5 .9 Pure 8.2 3.2 8.9 8.5 5.9 3.0 6.3 .8 Three-Fourths 7.4 1.6 10.3 5.1 1.8 7.5 5.6 1.0 Mulattoes 5.8 2.7 4.3 6.6 6.2 .9 4.4 .6 Quadroons .. 3.3 4.3 —7.8 4.3 —.5 .7 1.5 Boys Pure and Three-Fourths 5.4 4.7 6.0 12.3 12.6 3.7 7.4 1.2 Mulattoes and Quadroons 5.5 5.6 4.4 2.2 7.1 —1.3 3.9 .8 Girls Pure and Three-Fourths 7.7 2.2 9.8 7.0 2.7 4.0 5.6 .9 Mulattoes and Quadroons 5.8 2.8 4.3 4.4 5.8 .5 3.9 .5 TABLE XLVII. Completion Test — Difference Between Scores of the White and Each of the Four Classes of Colored Subjects, by Grades — Richmond (Minus signs indicate greater scores by the colored subjects.) Grades 5A 5B 6A 6B 7A 7B lA 2A 3A 4A Av.P.E. Pure 2.7 1.1 3.4 3.0 1.6 8.4 9.2 6.7 1.9 7.7 4.6 .7 Three-Fourths 1.4 1.7 4.8 9.0 4.6 9.6 5.5 3.0 2.5 6.9 4.8 .6 Mulattoes 2.8 3.1 2.7 4.3 3.2 8.4 3.8 2.5 4.0 3.0 3.8 .3 Quadroons —2.7 - -1.4 2.3 10.9 .8 2.0 1.2 5.2 2.3 .9 Pure and Three-Fourths 2.1 1.5 3.8 6.0 3.4 9.1 7.6 4.3 2.2 7.4 4.7 .6 Mulattoes and Quadroons 2.3 2.5 2.6 6.0 3.2 9.1 2.8 2.4 3.1 3.5 3.7 .4 COMPARISON OF SUB-CLASSES OF NEGROES. 109 TABLE XLVIII. Completion Test — Percentage of the Score of the White Obtained BY Each of the Four Classes of Colored Subjects, by Age and Sex — Richmond Ages 12 13 14 15 16 17 Av. P.E. Boys Pure 73 68 73 50 63 65 65.3 2.1 Three-Fourths 81 93 69 43 53 118 76.2 7.3 Mulattoes 71 62 73 88 77 104 79.2 3.8 Pure 61 81 64 70 80 90 74.3 3.2 Three-Fourths 65 91 59 82 94 75 77.7 3.8 Mulattoes 72 84 83 76 79 97 81.8 2.1 Quadroons .. 81 83 128 85 102 95.8 5.7 Boys Pure and Three-Fourths 72 75 71 47 59 88 68.7 3.5 Mulattoes and Quadroons 71 71 79 90 77 104 82.0 3.5 Pure and Three-Fourths 63 87 61 75 91 87 77.3 3.8 Mulattoes and Quadroons 72 84 83 84 80 98 83.5 1.8 TABLE XLIX. Completion Test — Percentage of the Score of the White Obtained BY Each of the Four Classes of Colored Subjects, by Grades — Richmond Grades 5A 5B 6A 6B 7A 7B lA 2A 3A 4A Av. P.E. Pure 79 93 81 86 92 68 66 78 94 77 81.4 2.0 Three-Fourths 89 89 76 59 76 63 80 90 92 80 79.4 2.3 Mulattoes 78 79 85 80 83 68 86 92 87 91 82.9 1.3 Quadroons 121 109 87 50 .. .. 97 93 96 85 92.2 4.2 Pure and Three-Fourths 84 90 79 73 82 65 72 86 93 78 80.2 1.7 Mulattoes and Quadroons 82 83 86 73 83 65 90 92 90 90 83.4 1.6 TABLE L. Completion Test — Percentage op Each op Two Classes of Colored Subjects Reaching or Exceeding the Average op the White, by Age and Sex and by Grades — Richmond Ages 12 13 14 15 16 17 Av. P.E. Bovs — Pure and Three-Fourths 17 18 36 43 19.0 4.9 Mulattoes and Quadroons 12 37 33 50 37 100 44.8 7.0 Girls — Pure and Three-Fourths 57 9 7 22 44 23.2 6.3 Mulattoes and Quadroons 25 30 28 31 25 50 31.5 2.0 Grades 5A 5B 6A 6B 7A 7B lA 2A 3A 4A Pure and Three-Fourths 25 39 14 12 33 10 29 57 40 25.9 3.6 Mulattoes and Quadroons 41 27 29 17 29 43 29 44 20 27.9 2.4 110 THE PSYCHOLOGY OF THE NEGRO. It is evident that the results from the completion test show the same general standing for the four classes of negroes that was brought out by the mixed relations test: the sub- jects with the greater amount of white blood were superior. And the size of the probable errors, particularly in the case of the two-division classification, renders the results reliable. The standing of the various classes of negroes in the mixed relations tests, I and II, and in the completion test may be averaged. When this is done the figures show that the pure negroes scored 69.2 per cent, as high as whites; that the three-fourths pure negroes scored 73.2 per cent, as high as whites; that the mulattoes scored 81.2 per cent, as high as whites; and that the quadroons obtained 91.8 per cent, of the white score. The pure and three-fourths pure negroes, combined, obtained 70.8 per cent, of the score of the white subjects; the mulattoes and quadroons, combined, obtained 83.6 per cent, of the score of the whites. This suggests that it is sometimes questioned as to how far external appearances are indicative of racial purity. It is held that a certain proportion of the offspring of mixed races will not show the blood characteristics which their heredity would seem to warrant. This may be true, and the Mendelian laws of inheritance may furnish a basis for such variation. But in the main, children conform closely to the character- istics of their parents in all respects, and such variations as occur will offset each other in any considerable number of persons, and leave the group type predominant. Certainly the results here obtained indicate that the correlation between skin color and racial purity is high. There was apparently nothing except native racial ability that could bring about the results found herein for the differ ent classes of negroes. It is possible, indeed, that the lighter negroes were of a better social class than the darker. But if this was true, it must be that the mixed bloods attained to a better social standing because of their greater capacity. For among negroes in general there are no considerable social dis- tinctions based on color. A colored person is a colored person, whether he be mulatto or negro, and all mingle together as one race. Pyle, in a quotation previously given, reports that of the negroes tested by him, those of better social class made higher scores. And it would not be surprising if it were found that mulattoes constituted the bulk of the superior COMPARISON OF SUB-CLASSES OF NEGROES. HI social class, and only a minority of the inferior class. It would be surprising, in view of the results found here, if it were otherwise. But any such distribution should be ascribed to the greater inherent capacity of negroes of lighter skin. No social or material gulf, such as that which separates the white and colored races in this country, is to be found among the negroes themselves. Such distinctions as there are, and however based, are very minor as compared with the great inter-racial distinction. And to a marked degree the general environment of colored children is the same for all. With no great variations, they all attend the same schools, live in the same neighborhoods, grow up in the same home surroundings, share the same cultural advantages, meet the same opportuni- ties, undergo the same experiences. They live almost in a world of their own, which is but one stratum or level of our mixed society. This is particularly true of the selected group of negroes found in the upper grades of the schools. And yet the differences, as revealed by the mental tests, between the pure negroes and the mulattoes, on the one hand, and the mu- lattoes and the whites, on the other, are not unlike in amount. The mulattoes who, with the pure negroes, live in the com- paratively uniform colored environment, which is so greatly different from the white environment, are yet almost as near to the whites as to the pure negroes when put to psychological tests. This fact becomes especially noteworthy when it is considered that the greater number of mulattoes are probably descended from an inferior element among the whites, and probably from an inferior element among the negroes also. Such considerations indicate that it is a native ability and not an acquired capacity that differentiates the mixed and the pure negroes, and that skin color is its outward sign. They also indicate that the tests used are primarily tests of native capacity, and that consequently the differences found between whites and negroes as a whole are innate differences. Comparative Variabilities In recent years a number of writers have called attention to the importance of differences in the variability of groups from their central tendencies. It is pointed out that two groups, as two races or two sexes, may be of the same aver- age capacity, but that if one of these groups is more variable than the other, certain significant differences will appear. 112 THE PSYCHOLOGY OF THE NEGRO. The more variable group will produce a greater number of individuals of very high capacity and also a greater number of very low capacity; within the less variable group there will be more conformity to the group type and less diver- gence toward the extremes. In the former case there will be a greater number of geniuses and of moral heroes than in the latter, and also a greater number of idiots and of moral de- generates. This will result in marked differences in the rela- tive attainments of the groups as wholes. Progress depends upon the few who stand out from the crowd and invent new and better ways of doing things. The few conceive and the many appropriate their conceptions, in all realms of human activity. In mechanical and material progress, in science, in art, in literature, in religion, in politics, it is the geniuses who make progress possible. And it follows that a variable group, other things being equal, will achieve a higher type of civilization than one which is less variable. In discussing this matter, Woodworth writes as follows: "The distribution of a trait is for some purposes more im- portant than the average. Let us suppose, for instance, that two groups were the same in their average mental ability, but that one group showed little variation, all of its mem- bers being much alike and of nearly the average intelligence, while the other group showed great variability, ranging be- tween the extremes of idiocy and genius. It is evident that the two groups, though equal on the average, would be very unequal in dealing with a situation which demanded great mental ability. One master mind could supply ideas for the guidance of the group, and his value would far outweigh the load of simpletons which the group must carry." ('10, p. 2). Thomdike ('10) writes to the same effect. He claims that men are more variable than women, and that on this account the greater part of the genius and the stupidity of the world is found among men. In discussing racial capacity, he says: "The comparison in variability is, as in the case of the sexes, of great practical importance. The ability of a hundred of its most gifted representatives often counts more for a na- tion's or a race's welfare than the ability of a million of its mediocrities." '10, pp. 53-54). While the importance of variability is thus recognized, there is as yet very little evidence bearing upon the relative variability of diffrent races. In summing up the psychologi- COMPARISON OF SUB-CLASSES OF NEGROES. 113 cal results available in 1910, Wood worth wrote as follows: "The dead level of intelligence, which is sometimes supposed to obtain among backward races, is not borne out by psycho- logical tests, since individual differences are abundantly found among all races, and, indeed, the variability of different groups seems, from these tests, to be about on a par." ('10, p. 15). Hrdlicka ('98) and Le Bon ('98), as quoted in Chapter I, claim that in physical traits whites are more variable than negroes. Strong ('13) states, as previously quoted, that the lighter-colored negroes among the 122 tested by her were more variable than those of darker color. Mayo, after dis- cussing the variabihty of whites and negroes, writes as follows concerning the groups studied by him: "In our own study of the two groups of high school pupils, however, the fact of greater racial mental variability is not at all pronounced, though the whites were slighly more variable. The average deviation of the white group from their mean scholastic at- tainment was 7, while that of the colored group was 6.5." ('13, p. 69). In addition to the somewhat inconclusive find- ings of these writers, there do not seem to be any reliable measurements bearing upon the problem. At present we have no good method of measuring the vari- ability of groups. Of course the actual variabihties may be compared in such terms as the average deviation or the probable error. But this will be misleading if the groups differ considerably in their standing in the capacity in ques- tion. A higher average standing generally implies a greater average deviation. If an average score of 30 has an average deviation of 10, an average score of 20 in the same trait will have a smaller deviation. The actual size of the deviations must therefore be considered in connection with the size of the average scores. In order to overcome this difficulty, Pearson has proposed that the deviations should be divided by the group averages before a comparison is made. But this procedure is open to the objection that it is not certain that deviations vary in proportion to the size of the central tendencies from which they are derived. There is reason to believe that they more nearly vary in proportion to the size of the square roots of the central tendencies. Accordingly Thorndike ('04) has suggested that actual variabilities should be divided by the 114 THE PSYCHOLOGY OF THE NEGRO. square roots of averages rather than by the averages them- selves. Since there is thus no single reliable measure of variability, our procedure in the following comparisons will be to set forth the average deviations, the Pearson coefficients, i.e., the deviations divided by the averages, and the Thorndike co- efficients, i.e., the deviations divided by the square roots of the averages. And since none of these methods is wholly free from objection, the average of all three is presented in each case as a combined coefficient. The variabilities are given in this manner for the various classes of colored subjects in terms of percentages of the variability of white subjects of the same ages or grades. The variability of the whites is therefore to be considered as 100 in all cases. The figures for the separate ages and grades are omitted for the sake of brevity, and only the aver- ages of the different ages and grades are presented. The actual deviations from which the figures are derived are shown in the tables in this and the foregoing chapters which set forth the scores. Only the Richmond pupils are treated in the comparisons. Table 51 shows the relative variabilities of the different classes of subjects in the age at which they reached the school grades studied. From the averages it appears that the pure negroes, the three-fourths negroes, the mulattoes and the quadroons were 66, 91, 87 and 74 per cent., respectively, as variable as the whites; that the pure and three-fourths ne- groes combined and the mulattoes and quadroons combined were, respectively, 83 and 95 per cent, as variable as the whites; and that the colored pupils of all classes combined TABLE LI. Ages — Variability of the Colored Subjects in Percentages of THE Variability of ' fhe White — Richmond- -By Grades Average Pearson Thorndike Deviation Coefficient Coefficient Av. Pure Negroes 68 65 66 66 Three-Fourths 92 90 91 91 Mulattoes 88 86 87 87 Quadroons 74 74 74 74 Pure and Three-Fourths 85 81 82 83 Mulattoes and Quadroons 96 94 95 95 AH colored 90 87 89 89 COMPARISON OF SUB-CLASSES OF NEGROES. 115 had 89 per cent, of the white variability. The average devia- tions, the Pearson and the Thorndike coefficients all give practically the same result: the negroes were less variable than the whites. The uniform behavior of the figures for age variability is in contrast with the more uncertain nature of those which show the variability of the scores in the tests. Tables 52-55 make the test comparisons. Relying upon the averages, it appears that the pure negro boys are 72, 106 and 63 per cent, as variable as the white boys in the Mixed Relations Test I, the Mixed Relations Test II and the completion test, respec- tively ; that the pure negro girls are 104, 98 and 84 per cent. as variable as the white girls in these respective tests; and that the pure negroes when compared by grades are 85, 85 and 89 per cent, as variable as the whites in these tests. The figures which show the relative variability of the three-fourths pure negroes in the three tests, in the order mentioned, are as follows: Boys— 69, 86 and 82 per cent.; Girls— 101, 88 and 56 per cent. ; Grades — 99, 110 and 85 per cent. The relative variability of the mulattoes in the three tests is: Boys — 90, 87 and 115 per cent.; Girls— 123, 104 and 102 per cent.; Grades — 109, 102 and 100 per cent. The percentages for the quadroons are: Girls — 118, 84 and 57; Grades — 69, 44 and 47. These figures are not subject to any sure interpreta- tion. But it appears to be very probable that the pure and the three-fourths pure negroes are less variable than whites, and that the quadroons are also less variable. The com- paratively small number of quadroons, however, may be a factor here. The mulattoes appear to have a variability as great as that of the whites. TABLE : LII. Mixed Relations Test- Percentages OF THE -Variability Variability OF OF the THE Colored Subjects White — Richmond IN Boys Pure Negroes Three-Fourths Mulattoes Test ] Average Deviation 54 58 79 [. Pearson Thorndike Coefficient Coefficient 92 71 81 68 101 90 Av. 72 69 90 Pure and Three-Fourths Mulattoes and Quadroons 65 89 106 107 82 97 84 98 All colored Girls Pure Negroes 90 85 124 125 106 103 107 104 116 THE PSYCHOLOGY OF THE NEGRO. Three-Fourths Mulattoes Quadroons Pure and Three-Fourths Mulattoes and Quadroons All colored Grades Pure Negroes Three-Fourths Mulattoes Quadroons Pure and Three-Fourths Mulattoes and Quadroons All Colored 89 114 Test II. Boys Pure Negroes 84 130 Three-Fourths 68 106 Mulattoes 78 96 Pure and Three-Fourths 85 135 Mulattoes and Quadroons 92 105 All Colored 101 128 Girls Pure Negroes 78 120 Three-Fourths 73 104 Mulattoes 92 117 Quadroons 81 87 Pure and Three-Fourths 83 120 Mulattoes and Quadroons 86 107 All Colored 89 114 Grades Pure Negroes 72 99 Three-Fourths 93 129 Mulattoes 91 115 Quadroons 40 48 Pure and Three-Fourths 85 117 Mulattoes and Quadroons 84 105 All Colored 87 115 Average Deviation Pearson Coefficient Thomdike Coefficient Av. 85 114 117 118 131 118 100 123 118 101 123 118 88 114 121 128 103 121 104 121 103 123 112 113 72 86 98 64 98 114 120 73 84 98 109 69 85 99 109 69 81 J 94 111 112 95 103 96 103 102 104 85 86 108 99 114 96 88 104 84 100 96 101 85 109 101 44 100 95 100 102 106 86 87 109 99 114 98 88 104 84 101 96 101 85 110 102 44 101 95 101 TABLE LIU. Completion Test — Variability op the Colored Subjects in Per- centages OP the Variability of the White — Richmond Boys Pure Negroes Three-Fourths Mulattoes Average Deviation 50 71 102 Pearson Coefficient 77 94 129 Thomdike Coefficient 62 82 114 Av. 63 82 115 Pure and Three-Fourths 68 98 82 83 COMPARISON OF SUB-CLASSES OF NEGROES. 117 Average Deviation Pearson Coefficient Thorndike Coefficient Av. Mulattoes and Quadroons 100 121 111 Ill All Colored Girls Pure Negroes Three-Fourths Mulattoes Quadroons 85 71 49 92 56 109 97 63 113 57 97 83 56 102 57 97 84 56 102 57 Pure and Three-Fourths Mulattoes and Quadroons 68 89 89 106 78 98 78 98 All Colored Grades Pure Negroes Three-Fourths Mulattoes Quadroons 71 80 75 91 45 89 98 95 109 49 81 89 84 100 47 80 89 85 100 47 Pure and Three-Fourths Mulattoes and Quadroons 82 84 102 100 93 93 92 92 All Colored 84 102 93 93 TABLE LIV. Maze Test — Variability of the Colored Subjects in Percentages OP the Variability of the White — Richmond Average Pearson Thorndike Deviation Coefficient Coefficient Av. Boys— All Colored Touches 87 126 104 106 Distance 91 109 100 100 Girls— All Colored Touches 105 174 136 138 Distance 121 153 136 137 Grades — All Colored Touches 82 118 100 100 Distance 91 105 100 99 TABLE LV. Cancellation Test — Variability of the Colored Subjects in Per- centages OF THE Variability of the White — Richmond Average Pearson Thorndike Deviation Coefficient Coefficient Av. Boys — All Colored Omissions 172 89 122 128 Cancellations 103 105 104 104 Girls — All Colored Omissions 153 113 133 133 Cancellations 127 118 120 120 Grades — All Colored Omissions 167 106 133 135 Cancellations 103 94 98 98 118 THE PSYCHOLOGY OF THE NEGRO. The combined pure and three-fourths pure negroes, when compared with whites, have relative variabilities as follows in the Mixed Relations Test I, the Mixed Relations Test II and the completion test, respectively: Boys — 84, 109 and 83 per cent.; Girls — 104, 101 and 78 per cent.; Grades — 96, 101 and 92 per cent. The figures for the combined mulattoes and quadroons show that they have the following relative varia- bilities in the three tests: Boys — 98, 99 and 111 per cent.; Girls— 121, 96 and 98 per cent. ; Grades— 103, 95 and 92 per cent. It is thus probable that the mulattoes and quadroons are slightly more variable than the pure and three-fourths pure negroes, but here again the results are not certain. Both classifications appear to have not far from the white varia- bility. When all classes of negroes are grouped together and com- pared with whites, they show the following relative variabili- ties in the three tests mentioned: Boys — 107, 114 and 97 per cent. ; Girls — 113, 101 and 80 per cent. ; Grades — 102, 101 and 93 per cent. The colored subjects as a whole had a greater rather than a less variability than the white. This last statement is reinforced by the figures from the maze and cancellation tests. In the maze test, the colored boys were 106 and 100 per cent, as variable as the white boys in touches and distance respectively; the colored girls were 138 and 137 per cent, as variable as the white girls in touches and distance; the colored grades were 100 and 99 per cent. as variable as the white in touches and distance. In the can- cellation test, the figures showing the relative colored varia- bility in omissions and cancellations, respectively, were: Boys — 128 and 104 per cent.; Girls — 133 and 120 per cent.; Grades — 135 and 98 per cent. The negroes in these tests were more variable than the whites, all classes of colored sub- jects being grouped together. In addition to these figures from Tables 52-55, attention should be called to those in Tables 39-43 and 46-50 as indica^ tions of the relative variability of the different sub-classes of negroes. These tables give the probable errors of the scores made by the various classes of negroes when compared with whites. The probable errors show a marked uniformity in being smaller for the darker^colored negroes and larger for those of lighter color. That is, they show that in the case of the darker negroes the separate age and grade groups va- ried less with respect to their own average difference from COMPARISON OF SUB-CLASSES OF NEGROES. 119 corresponding white age and grade groups. Now this greater constancy among the differences for the separate ages and grades of darker negroes might have been due to chance if it had occurred only infrequently. For example, if the scores of four white grades are 40, 50, 40 and 50, and the scores of four corresponding colored grades are 20, 30, 20 and 30, the differences between these grades will be 20, 20, 20 and 20, and the probable error will be zero. But if the four corresponding colored grades have scores of 30, 20, 30 and 20, the differences will be 10, 30, 10 and 30, and the probable error will be 5. The actual scores and the actual average differences are the same in the two cases, but a chance arrangement in the correspon- dence of the scores alters the probable errors. And if the difference from the whites of first the lighter and then the darker negroes had the smaller probable error, we could at- tach no significance to the fact. But the smaller probable error in the case of the darker-colored negroes occurs with a uniformity which cannot be ascribed to chance. And we must suppose that it is due to a greater constancy in the scores of the individuals within the darker age and grade groups themselves. This greater constancy of the individual scores would be reflected in the greater constancy of the separate group scores and hence in the smaller probable error of such scores. And consequently we have a further indication of the smaller variability of the negroes of comparatively pure blood. On the whole, it appears that the pure negroes, the three- fourths pure negroes and the quadroons varied less than did the whites ; that the mulattoes did not differ from the whites in variability ; that the combined pure and three-fourths pure negroes and the combined mulattoes and quadroons had not far from white variability, and that the latter class varied slightly more than did the former; that the colored subjects of all classes together had a greater variability than the white. But these conclusions are only approximately certain. The average coefficients upon which they are based are not uni- form in their import, and the separate coefficients taken by themselves are not more so. If we consider only the actual variabilities, the white are clearly more variable than the col- ored subjects, except in the cancellation test, and the sub- classes of negroes, particularly those of darker color, are much less variable than the negroes as a whole. It is in- teresting to note that the colored girls had generally a greater 120 THE PSYCHOLOGY OF THE NEGRO. variability than the colored boys when compared with whites. Figures 25-28 are inserted to show graphically the relative distribution of the two races in one test — ^the completion. This distribution is fairly typical of that found in the other tests. All classes of negroes are included. In the graphs the boys and girls are treated separately, but the elementary pupils are grouped together, as are the high school pupils. Score O r n /6 ^o 2v 7r sx 36 ^/o ^y ^t sc Fig. 25. Completion Test — Distribution of the Scores of White and Colored Subjects — Grammar-Grade Boys — Richmond. The solid and the broken lines indicate the scores of the white and the colored subjects, respectively. Pet. 20 Score. J: V ^ /2 i — i ^ Fig. 26. Completion Test — Distribution of the Scores of White and Colored Subjects — Grammar-Grade Girls — Richmond. COMPARISON OF SUB-CLASSES OF NEGROES. 121 Pet. IX % "~ '*'*i 1 _r^— • 1 1 1 I _J 1 1 1 • ■"^ ^-. Score O V ^ 12 JL 20 7i 28- 32 3C "/O Hf "/S- Cro Fig. 27. Completion Test — Distribution of the Scores of White and Colored Subjects — High School Boys — Richmond. Ptt. 20 fC a S H u^a Scon O V ? 12 U 20 2"/ 2» 32 3^ V^ VV ^-^ 50 Fig. 28. Completion Test — Distribution of the Scores of White and Colored Subjects — High School Girls — Richmond. This procedure is permissible on the ground that within either the elementary or the high school grades tested the pupils do not differ significantly from age to age, as was shown in Chapter III. The number of pupils obtaining a given score was in all cases reduced to a percentage of the total number in the distribution. From the graphs it appears that the white elementary boys and the white elementary girls are probably more variable than the colored, but that the colored subjects, particularly the boys, are probably the more variable in the high school. But no certain conclusion as to racial variability can be reached. It appears more certain that the boys are more variable than the girls. Incidentally, attention may be called to the comparison of racial ability revealed by the graphs. They make plain the extensive overlapping of the scores of the whites and negroes. Even where there is a great differ- 122 THE PSYCHOLOGY OF THE NEGRO. ence between the average scores of the two races and where only a small percentage of the negroes reaches or exceeds the average of the whites, the overlapping is great and should not be overlooked. It may finally be remarked with regard to the relative vari- ability of whites and negroes, that it would not be at all sur- prising if groups of so-called negroes were definitely shown to be more variable than comparable groups of whites. For the groups that are generally called negroes are composed of in- dividuals ranging from pure negroes to persons almost white, and it would be reasonable to suppose that such groups would vary more than would homogeneous white groups. They are not composed of one race but of two. Of course it may be that the variability of the negro race falls entirely within the extremes of variability of the white, but this would not offset the greater average deviation that would be caused by the relative tendency to bi-modality, or at least to flatness of distribution, in a mixed negro and mulatto group. This consideration should be taken into account in all studies of the variability of whites and negroes. It would be interest- ing and useful to know whether colored people, negroes and mulattoes together, vary as much as do whites. But this information would not inform us as to the relative variability of the white and negro races. CHAPTER V. CONCLUSION By way of summary of the various considerations which have come to light in this study, we may say that the average performance of the colored population of this country in such intellectual work as that represented by the tests of higher capacity, appears to be only about three-fourths as efficient as the performance of whites of the name amount of trainings' It is probable, indeed, that this estimate is too high rather than too low. The groups of whites and negroes studied are not typical of the white and colored populations in general; their ability is undoubtedly considerably above the average. But the negroes were probably farther above their racial av- erage than were the whites. If one were to test the capacity of the unselected masses of negroes, with their much smaller per- centage of white blood, and make a comparison with unselect- ed masses of whites, the results would almost certainly reveal greater racial differences than those found herein. All of the experimental work which has been done has pointed to the same general conclusion. The bulk of it has shown a greater racial difference than that found in Rich- mond, and has been more comparable in its results with the findings recorded for Fredicksburg and Newport News. The opinions of the great majority of those who have come into contact with the negro, and the views of nearly all of those who have studied the question from standpoints other than experimental, are in substantial agreement with the quantita- tive evidence. In the present state of the advancement of science it does not seem possible to make adequate tests of those vastly im- portant qualities which are included in the feeling and dy- namic, rather than in the intellectual, side of mental life. It is the common opinion that the negro differs more from the white in such traits than in intellect proper. His emotions are generally believed to be strong and volatile in their mani- 123 124 THE PSYCHOLOGY OF THE NEGRO. f estations ; whether this is due to their intrinsic nature or to a lack of restraint, is an untouched problem. Instability of character is ascribed to the negro, involving a lack of fore- sight, an improvidence, a lack of persistence, small power of serious initiative, a tendency to be content with immediate satisfactions, deficient ambition. But the evidence that such characteristics constitute a true racial difference cannot be called conclusive, and the psychological causes underlying them have not been adequately investigated. Along with high emotionality and instability of character, defective mo- rality is held to be a negro characteristic. This is as subject to debate as are the other qualities, though it is apparently supported by social statistics.* It may be that the total cir- cumstances of his life are such as would lead to immorality even were the negro possessed of the psychic nature of the white man. On the other hand, while it is impossible to arrive at an exact knowledge of the relative amounts of such important but intangible traits in the two races, it must be said that the evidence of experience and observation cannot be disregarded. Such evidence is often wholly unscientific and worthless, but not always so. Strong and changing emotions, an improvi- dent character and a tendency to immoral conduct are not unallied. They are all rooted in uncontrolled impulse. And a factor which may tend to produce all three is a deficient de- velopment of the more purely intellectual capacities. Where the implications of ideas are not apprehended, where thought is not lively and fertile, where meanings and consequences are not grasped, the need for the control of impulse will not be felt. And the demonstrable deficiency of the negro in in- tellectual traits may involve the dynamic deficiencies which common opinion claims to exist. The available evidence indicates that in the so-called lower traits there is no great difference between the negro and the white. In motor capacity there is probably no appreciable *Statistics concerning the sexual immorality, as indicated by ille- gitimate births, of whites and negroes in the District of Columbia have recently been published by Ottenberg ('15). In 1912 and 1913 the total of all births reported to the Health Department was 13,910, of which number 1374, or approximately 10 per cent, were illegitimate. There were four times as many illegitimate births of colored as of white chil- dren reported, and yet the colored population was only about one-half as large as the white. Statistics showing the very much larger per- centage of negroes than of whites convicted of crime are too well known to require quotation here. CONCLUSION. 125 racial difference. In sense capacity, in perceptive and dis- criminative ability, there is likewise a practical equality. It is in the central elaborative powers upon which thought more directly depends that differences exist, not in the simpler re- ceptive and discharging functions. %t seems as though the white type has attained a level of higher development, based upon the common elementary capacities, which the negro has not reached to the same degree".; From the nature of the men- tal differences, one would infer that such neural differences as may be found will probably be mainly in the constitution of the cortical neurones, rather than elsewhere in the nervous system. While the intellectual performance of the general colored population is approximately 75 per cent, as efficient as that of whites, this figure is not true for different classes of negroes. It is probably correct to say that pure negroes, negroes three- fourths pure, mulattoes and quadroons have, roughly, 60, 70, 80 and 90 per cent., respectively, of white intellectual effici- ency. If it were possible to distinguish these four classes of negroes so accurately as to avoid overlapping, it is probable that the differences revealed by tests would be greater rather than less than those indicated by the figures. The educational significance of the available facts is difficult to determine. The negro's intellectual deficiency is regis- tered in the retardation percentages of the schools as well as in mental tests. /'And in view of all the evidence it does not seem possible to Vaise the scholastic attainment of the negro to an equality with that of the whitei It is probable that no expenditure of time or of money would accomplish this end, since education cannot create mental power, but can only de- velop that which is innate. The movement toward industrial education for the negro finds sanction in the studies of his psychology. Without great ability in the processes of abstract thought, the negro is yet very capable in the sensory and motor powers which are involved in manual work. And economy would indicate that training should be concentrated upon those capacities which promise the best return for the educative effort ex- pended. Social conditions, of course, have been the main in- centive to the growth of industrial education among negroes, 126 THE PSYCHOLOGY OF THE NEGRO. and in themselves they are sufficient reason for emphasiz- ing an intensely practical training. But the mental nature of the negro gives reason for believing that this sort of edu- cation is the only one which will avoid great waste. Dimin- ishing educational returns will be more serious in the intel- lectual than in the industrial education of the negro. There is need of experiment to determine the relative ability of colored and white persons in the intelligent handling of concrete materials. All of the experiments so far undertaken have dealt with thought material as represented by words rather than by objects. Tests which involve mechanical ma- nipulations have not been tried. It is possible that reasoning based upon objects present to sense may not correlate highly with reasoning based upon the mental representations of ob- jects, and that the negro may therefore more nearly approach the white in the former than he does in the latter sort of thinking. The writer expects to undertake a series of experi- ments upon the comparative intellectual ability of the two races in mechanical tests, and if the difference between them is less than that revealed by tests of a more literary nature, additional sanction will be given to the reasonableness of in- dustrial education. But while it thus appears that for the colored population as a whole a manual is more practicable than a literary educa- tion, it must not be overlooked that there are individual col- ored persons of great ability. The widely held doctrine that the negro's mental growth comes to a comparative standstill at adolescence does not find corroboration in the results of tests. The groups so far tested, indeed, show that after adolescence the negro more nearly approachs the white than before. This is probably due to the fact that the adoles- cent negroes tested are a more closely selected group than those who have not reached adolescence. The adolescent negroes in the schools have more white blood in them. And racial differences at adolescence may exist in the feeling and dynamic sides of mental life, which have not been tested. If there are such differences they will most likely appear just here. But so far as has been demonstrated, the negro's intel- lectual development proceeds as rapidly after puberty as does that of the white. Then, too, the variability of the negroes and the overlapping of ability in the two races, make it neces- sary to expect very able colored persons to be found in every CONCLUSION. 127 large group. In the main, the most capable colored individ- uals will be mulattoes, although there are fewer mulattoes than pure negroes. ' v^ince comparisons between races are frequently made in terms of the number of eminent men produced by each, it is interesting to make a rough computation as to the number of pure negroes, mulattoes and whites that may be expected to demonstrate great ability in the United States. This com- putation lays no claim to other than approximate validity. It is based upon the "law of deviation from an average" as employed by Francis Galton ('92, p. 22 ff.), and in making it the variability of each class of the population is considered as being the same. Of course, if the variability of pure negroes is less than that of whites, as is probable, the number of very able negroes will be less than is indicated by the following figures. We may take as our standard of eminence the one chosen by Galton, viz., attainment so great that it is reached by only one man in 4300. In each million men there are 248 persons of this standing. Now there are approximately 18,000,000 white men, 1,600,000 pure negro men and 400,000 men who are mulattoes in this country. Consequently, if all three classes have the same ability, there will be 4464 eminent white men, 397 eminent pure negroes and 99 eminent mulat- toes. But if we assume that pure negroes average 75 per cent, of white ability and that mulattoes average 87.5 per cent, of white ability, we find the following situation growing out of the law of deviation from an average.* In a million of each class of men, there will be 248 whites, 15 mulattoes and 1 l^ure negro who will attain the above-mentioned degree of eminence. Considering the number of these three classes *To carry out the computation based upon this assumption it is neces- sary to postulate a zero point of ability, and the capacity of the lowest idiot is taken as this zero point. That this may be done without violence to the facts is indicated by the descriptions of idiocy given in works on mental deficiency. Thus Tredgold ('08) writes of profound idiots that their brief existence may almost be called vegetative. They are devoid of instincts, they lie huddled in an ante-natal posture, food must be placed in their mouths, and their life activities hardly extend beyond respiration, assimilation and excretion. "They have eyes, but they see not; ears, but they hear not; they have no intelligence and no conscious- ness of pleasure or pain; in fact, their mental state is one entire nega- tion." ('08, p. 171). 128 THE PSYCHOLOGY OF THE NEGRO. in the total population, there will be 4464 eminent whites, 6 eminent mulattoes and 2 eminent negroes in the United States. These figures are suggestive. If we take it that there are 4464 eminent white men in America, there are certainly not 397 pure negroes and 99 mulattoes of the same degree of eminence. There are more nearly 6 mulattoes and 2 pure negroes to 4464 eminent whites. Definite figures are not ob- tainable, but such lists of men of achievement as have been compiled accord with the latter set of figures far more closely than with the former. Of course it may be held that social conditions make it im- possible for colored ability to assert itself. There may be potentially eminent men among the negroes who are not able to attain their commensurate achievement on account of en- vironmental conditions. On the other hand, it may be said that the best opinion, as that of Galton, holds that eminence is independent of circumstance; that innate power can be neither crushed nor created by adverse or favorable influences. And it may be further contended that ability among negroes is all the more readily recognized just because of their gen- erally low level of racial attainment. A man of mark among them stands out becauuse of his rarity, and his opportunities are increased because of this recognition. As an indication of the greater ability of mulattoes than of pure negroes, it may be remarked that such lists as we have of colored leaders, e.g., those quoted in Chapters I and IV, show a larger proportion of men of mixed than of unmixed blood. And this despite the fact that there are probably four times as many pure negroes as mulattoes in the country. Although the available facts are very few and inexact, such as they are they serve to justify rather than to controvert the deductions from the law of variability. And in so far as the deductions are borne out, the assumed racial inequality upon which they are based is confirmed. There are few more controversial subjects than that of the outlook for the negro race in America, and it is not within the province of this monograph to attempt a discussion of the topic. But it may not be out of place to mention certain considerations that have presented themselves. Conclusions concerning the negro's possibilities in this country are fre- CONCLUSION. 129 quently drawn from a study of the various small negro re- publics, such as Haiti, Santo Domingo and Liberia, and opin- ions so arrived at are not without their value. Yet there are differences between the position of the American negro and that of the negro in the isolated states in question. It should be noticed that the number of American negroes is larger than the number in any of the negro republics. Progress de- pends upon the sv V ' %. :o^'^\ %^^' &^ ^ % i^ /c^f'^,;^ %.o^ •f V -^ . <^^ I/ / A Z cP^.^r^»V 0^ ^ ^' * '> / ^r "^ Ho. .^^ > %<^' :^i cP\-l"°- ^. '%<^^ ,u> 9^ ', SgliP^ . '^ . ^ "n Willis ."^ k4^ ^ "' ©lis / 1 ■C.P '>*,. » w- ' .^^^ -^'^^ f. # * .^ "^-.,^^' ;o Av V '"/^^^^ eP' -^^ '"'^c '^e^ ^ rO^ ^^'"^ '^r^ '^ cP '-^ 1 ■» , '/i. '-^' ^ o.^' '^^>««.V^^^ ^^^°-X "" V>-' *<> :^ %^''^7rs^\sf" , 95,' .>V,^;/%^<^ *'.*'^|':X ^^^ '• c .^' ^ ' ft * "> ^v- x^ r% , , .^^^ . < ■r A-v •> ^ '' / ^ . s -< AS* <' ' i c ^ "" A*^ ^ '' / . . s *• A^ ' ^ ^ ■^0^