great changes in a few cases is shown by the fact that 53 per cent. of the defectives improved and only 43 per cent. of the other children Cases Per cent. of Median Im- No. of Per cent. of Median Im- No. o Cases In the intelligence tests the greater improvement is with the ordinary children although there is decided improvement shown by the feeble-minded. The average of the median of improvement in the four intelligence tests for normal children is 4.3, while for the defectives it is +0.8. Looking at the whole series of tests, the contrast between the defective children and ordinary children is not so marked as might be expected; for in one half of the tests the former improved more than the latter and in the other half the reverse is true. In only two of the measurements do the defectives show no improvement whatever, while the ordinary children show the same lack in three of the tests. If the total amount of improvement in all the tests is considered the defectives have the advantage, 25.5 to 21.7. But this result is due largely to one measurement, in which the defectives have a mark of +5.0 and the ordinary children -9.5. Discarding this one test, the greater improvement is found to be with the ordinary children, 31.2 to 20.5. Comparing the improvement of the defective children in one class of test with their improvement in others, we find that there is least. improvement in the results of the intelligence tests, with an average improvement of 1.8; those of memory rank second with a mark of 2.2, while the greatest improvement is in the tests of maturity, the average being 2.4. These figures show that the feeble-minded improve most in those mental traits in which they are most like the ordinary child and least in those traits in which they are most unlike. In all this matter of improvement the case for the defectives may not be quite so good as it seems to be, for the standards which they first attained in the tests of maturity, memory and intelligence were far below those of ordinary children. For instance, in the first measurements only 1 per cent. of the defectives exceeded the median mark for school children in the maturity tests; 7 per cent. exceeded it in the tests of memory and 0 per cent. exceeded it in the tests of intelligence. With this state of things of course there was much greater room for improvement in the case of the defectives than there was for the public school children. Hence a unit of improvement for any of the former may have quite a different value from the same amount of improvement as found among the latter. The obvious objection to these results as stated is that the comparison made is not a fair one, for nine-year-old normals have been compared with defectives of all ages from 7 to 17. However, the unfairness has probably been toward the defectives rather than toward the ordinary children, for the gain between nine and ten years of age is usually greater than the improvement between any other two years later on. This is especially true if the improvement between nine and ten years is compared with that between thirteen and fourteen. As it happens there are more of the defectives of thirteen and fourteen than of any other two years and there are only six nine years old or under. The only right way to compare these two classes of children would be to compare equal numbers of the same age. As I had not the data to do this, my best way was (1) to get from the standards found in Tables VI. to XXIII. (see § 9) the amount of improvement shown in each test by normal children from year to year and (2) to weight each figure so obtained by the number of cases of defectives for that age. For example, in the A test the average number marked at eight years old was 30, and at nine years old 32.6: the improvement in a year's time would be +2.6. There are four of the defective children between eight and nine, hence +2.6 must be weighted by 4. Going through for each test in this way, we get a series of measurements of improvement comparable with those obtained from the defectives. The medians may then be reckoned from these measurements. The comparative records so obtained may be seen in the following table. The results as presented in this table are rather startling, giving as they do a total improvement mark exclusive of the measurements of height and weight of 24.1 to the defectives and one of only 8.7 to the same number of ordinary children of the same ages. However, there are some sources of error which reduce the validity of these figures. In the first place, the usual fallacy of selection is present; for children of a certain age are tested and then other children a year older are subjected to the same tests, with the idea that so we may find out what the first children will do a year later. In the second place, when the defectives were measured the second time the same tests were used, hence it was a second trial and in some cases a third trial with the same tests, whereas in following the method stated above with the public school children the tests were only taken once. That a second and third trial do give an undue advantage, at least in some of the tests, the figures in Table XXXV. show. In the third place, the units compared are not the same, for the level reached by some of the ordinary children in the first series of tests was very much higher than that reached by the defectives, hence the possibility of improvement in one case is much less than in the other. A mark of 1 improvement in the genus-species test for the public school children may be a higher mark than the 1.6 gained by the defectives. Is passing from 96 per cent. to 97 per cent. a less gain than passing from 70 per cent. to 75 per cent.? We do not know. These sources of error make the figures stated unreliable so far as comparative value is concerned. They must serve. simply as suggestions. Any one who can obtain records from ordinary children for two successive years in the same way that I have from the defectives, will then have the data to furnish a reliable answer to this problem. However valueless these figures may be from a comparative standpoint, they have a very definite intrinsic value. They show definitely that the feeble-minded do improve from year to year and that their improvement is no mean one. The results so far might mean that those defectives most like ordinary people go on improving and that those farthest from normal grow more and more unlike, resulting in a divergent series. To throw some light on this point the whole number of defectives from the Waverley Institution was divided into quarters according to their first marks in the intelligence tests. The first quarter was composed of those students having the highest marks, the second of those having the next highest, and so on in order to those possessing the lowest marks. The improvement in these same tests of each group was then found. The average improvement of the highest quarter was .8; of the second, 1.4; of the third, 1.7; and of the lowest, 1.0. Those defectives most like ordinary people have improved least. The lower half of the class has improved more than the upper half, the averages standing 13.5 to 11. The results are still more striking if one takes simply the extremes and considers the improvement of the best five in the group as compared with that of the worst. The average improvement of the best five is .6, while that of the worst is 1.9-three times as great according to the actual figures. Here, as before, it must be borne in mind that the units of improvement are not the same, consequently the figures measuring the two are not actually comparable. This, however, does not alter the fact that the defectives fairly low in the scale do improve decidedly in a year's time -that the improvement is not confined to those who most closely approach the ordinary. The relative dependence of physical and mental growth has always attracted considerable attention from psychologists. The six defectives who for the past year grew at the most rapid rate were picked out and also the six whose rate of growth was slowest. The average improvements of these two groups in the intelligence tests were compared. The results are 1.6 for the former and 2.1 for the latter. Those who grew most slowly have the greater increase in intellectual ability. Limiting this comparison to children of the same age, and so taking the twelve-year-olds whose rate of growth was most rapid and comparing with them in improvement in intelligence those whose rate of growth was slowest, the results are more striking-.7 for the former, 2.5 for the latter. The general conclusions-which of course are only tentative-of this study of improvement are then: (1) That among mental defectives a decided improvement in mental ability may be looked for after the lapse of a year, in some directions even exceeding that shown by ordinary school children. (2) That the greatest improvement is not confined to those defectives most like ordinary individuals. (3) That the improvement is not equal in all directions, but that some mental functions improve more rapidly and to a greater extent than others and that even the functions we designate as intellectual show a marked improvement. One or two measurements were taken which, although they do not bear on this question of improvement, are interesting from an historical point of view. When the temperature was taken the first time it was found that the defectives had, on the average, a lower temperature than the ordinary children. This result was corroborated by the second measurement. Another interesting fact was noted, namely, that in 72 per cent. of the cases the temperature was lower than at the first previous measurement, although at the close of an hour's mental work only 44 per cent. of the children had a lower temperature than that taken under the same conditions the previous year. Whether this fact is significant or what the cause for it may be, there is as yet no means of determining. The only difference in the conditions in which the measurements were made was a difference in season, as the first tests were made at the beginning of the winter and the second at the close. Since physicians in general do not note any difference in the temperature of normal individuals due to the change of seasons, it would be rash to suggest this as a cause. The eyes of 44 of the defectives were tested by the use of Snellen's E test. The results are as follows: 48 per cent. could read the lette correctly at the distance specified; 52 per cent. could not. In giving this same test at the St. Louis Exhibition, Woodworth and Brune found that the average ability of whites under twenty-five years o age could be represented by 1.68. As only 48 per cent. of the de fectives could be given a grading of 1.00 the difference is ver marked. Whether these results are comparable may be questione but certainly the results are suggestive. It has always been claime that the sense organs of the defectives are very poor. In fact fixed is this notion that it may cause surprise that these figures not show the conditions to be worse. In testing the sensitiveness of the skin of 29 defectives with æsthesiometer the points of which were 20 mm. apart, the meth |