ence can be overcome by training, for a number of these Chinese students who have had considerable training in English can read Chinese as well or better when the characters are horizontally arranged. Our facts can be wholly explained in terms of habit and training. The concept of habit nicely accounts for the fact that the majority of the individuals did better with the vertical arrangement of Chinese characters and with the horizontal arrangement of the English materials; while the individual deviations from the normal preferences can be explained in terms of conflicting habits. Those individuals with the least amount of training in English made better records with the vertical arrangement of Chinese characters and with the horizontal arrangement of English letters and Arabic numerals. We have on the one hand the thoroughly established habit of reading Chinese characters with the vertical arrangement with but little opportunity for disintegration through disuse, and on the other the newly developed habit of reading English in the horizontal direction that is incapable of functioning with any high degree of speed and precision. Under these conditions, the two habits function with a minimal amount of conflict and interference. At the other extreme, those individuals with the maximum of training in English do as well or better with the horizontal arrangement of Chinese characters. In this case the habit of reading in the horizontal direction has become thoroughly established, and the habit of reading in the vertical direction has become somewhat weakened by disuse. In other words, the habit of reading English when thoroughly established becomes somewhat generalized and interferes with, or tends to suppress, the older habit of reading in the vertical direction. Finally, those individuals with a moderate degree of training in English experience a certain amount of difficulty with the vertical arrangement of Chinese characters as compared with those with the lesser training in English, and the facts, though not entirely convincing, indicate that these same individuals can not read English when horizontally arranged with the same relative facility as can those with less training. Under these conditions of relative strength, the two habits conflict and mutually interfere with each other. This hypothesis is at least suggestive from the standpoint of the study of the mutual relations of two antagonistic habits. A COMPARATIVE STUDY OF STYLUS MAZE LEARNING BY BLIND AND BY HELEN L. KOCH AND JENNETTE UFKESS The purpose of the study was to check with more adequate controls with respect to age, intelligence, and number of subjects some incidental observations made a few years ago by one of the authors upon the maze-learning of the blind. These observations, which are quoted by Dr. H. A. Carr in "The influence of visual guidance in maze learning," 1 indicate that the usual maze presents to the blind more difficulties than to those with normal vision. This result is distinctly contrary to a priori judgment, for a stylus-maze was employed which all of the subjects mastered without the aid of vision; and we are likely to argue in this case that the blind, who make so many of their spatial adjustments on a tactual-kinæstheticmotor plane, will find little that is novel or difficult in the maze problem; whereas the normal subjects, deprived of the visual cues upon which they are wont heavily to depend, will be considerably disturbed. We tend to overlook, however, the possibility that the mere fact of visual experience may provide a more adequate system of spatial symbols than does the sensorially more restricted experience of the blind. These action-patterns developed through vision may function, in other words, to overbalance any handicap to the normal subject in the maze resulting from his usual modes of reacting. The blind subjects, 19 in number, who served in the present investigation constituted the adolescent group in the Texas School for the Blind at Austin. The ages, as indicated in Table III., range from 14 to 26 years. Some of the subjects have not been totally blind from birth, and some even now possess what is known as 'shadow vision.' A description of the visual capacities of each subject is given in Table I. 1 This JOURNAL, 1921, 4, 399–417. TABLE I YEARS AND TYPE OF VISUAL EXPERIENCE BEFORE ONSET OF BLINDNESS Subj. IB-6B.. IPB. 2PB. 3PB. 4PB-8PB. 9PB. 10PB. 11PB. 12PB. 13PB. Description of Visua! Experience All totally blind from birth. Detail vision for 6 years, then total blindness. Shadow vision for 14 years, then total blindness. Detail vision for 7, 12, 3, 9, 5 years (resp.), then total blindness. Detail vision for 8 years, then total blindness. . Shadows and gross details for 17 years, then total blindness. Those individuals who have not always been deprived of their sight are valuable for our analysis, since not only do they furnish us an interesting intermediate to the extreme groupsi.e., those with average vision and those with complete lifelong blindness-but they also enable us roughly to gauge the effect of varying degrees of visual experience. The 20 normal subjects were selected from at least 100 senior-high-school and university students, as those who paralleled most successfully in age and intelligence the blind group. All but 5 of the 20 are high-school students. The. age range of the normal subjects is from 14 to 32 years. 3 An approximate pairing of the two groups on the basis of age and intelligence was deemed advisable, first, because of results like those of Hunter and of Warden, which indicate a correlation of moderate degree between skill in the mastering of mazes and performance in the intelligence-test; and, secondly, because the blind group, taken en masse, may perhaps be suffering some intellectual handicap as a result of the general physical and nervous disabilities produced by the same factors which mediated their visual defects. Table II. suggests that the caution was justified for, in the case of both blind and normal subjects, homogeneous and small though the groups were, a consistent, medium-sized, negative correlation 2 Hunter, W. S., Correlation studies with the maze in rats and humans, Comp. Psychol. Monog., 1922-23, I, no. 1, pt. ii, 37-56. 3 Warden, C. J., The relative economy of various modes of attack in the mastery of the stylus maze, This JOURNAL, 1924, 7, 243–276. * The Pearson product-moment method of computing correlation was used throughout this study. TABLE II CORRELATIONS OF AGE AND INTELLIGENCE-TEST SCORE WITH EACH OTHER AND WITH MEASURES OF MAZE PERFORMANCE. BLIND AND NORMAL GROUPS is shown between intelligence-test scores and the difficulty experienced in mastering the maze, as represented by the trials or time consumed or by the errors accumulated. The correlations tend to be slightly greater in the case of the blind group than they are in the normal, probably because the former is somewhat less homogeneous than the latter. Our results, then, in general bear out the results of Hunter and of Warden. In our blind group, age and maze success do not seem so closely related as do intelligence and the latter, the subjects being practically adult. The positive correlations of moderate size in the case of the normal group between age and the criteria of skill in maze performance, or rather lack of skill as our data are presented, are, doubtless, the result of the method of selecting these subjects. Most of them were juniors taking chemistry in the summer session of the senior-high-school. Since attendance at summer sessions is usually conditioned either by dullness or failure during the year or by unusual interest, success, ambition, etc., we should expect the younger subjects to be generally better equipped from the point of view of intellect than the older ones. The slight negative correlation between age and intelligence supports this hypothesis. Since no adequate intelligence tests for the blind adolescents or adults were available, the authors found it necessary to adapt to their needs one of the standard group-tests. They are fully aware that in their adaptation the character of the test has been altered and that probably the claims made for the tool as a measure of intelligence in its original standardized form are no longer entirely applicable. Certainly the speed factor was to a large extent eliminated, since the problems were necessarily presented orally, and the subjects at least had (so to say) a chance at every task. It is well to note, also, that while both blind and normal subjects use their |