| I.Q. | Per cent of Tonsil Group | Per Cent of Normal Group |
| Below 70 | 4·1 | 2·1 |
| Below 90 | 32·0 | 29·6 |
| Above 110 | 10·9 | 11·1 |
| Above 120 | 3·7 | 3·0 |
| Above 130 | 1·2 | 0 |
In other words, in the percentage of cases below normal intelligence, the tonsil group exceeds by 2·4 per cent. The percentage of defective cases is also slightly greater in the tonsil group—the difference here being 2 per cent. The normal group has a negligible predominance of bright cases,—only two-tenths of one per cent difference, while with the very superior cases, the tonsil group again exceeds,—by 1·2 per cent. The per cent of the tonsil group which reaches or exceeds the median of the normal is 49 per cent.
These figures seem to indicate remarkable similarity between the two groups considered. The two distributions are almost identical. While the slight predominance of cases below normal mentality in the tonsil group may indicate a very feeble tendency toward coincidence of tonsillar defect and mental dullness, it does not seem large enough to be at all significant. This is especially true when we consider that the tonsil group exceeds in superior children. If we allow the preceding contention of coincidence between dullness and tonsils, must we not argue here in the same manner for a tendency toward coincidence of superiority and tonsils?
The chief source of error in this part of the study is the fact that the throat examinations were not conducted by the same person throughout the investigation. For this reason there must have been some slight disagreement as to what should constitute a reportable case. In the event, then, of a positive relationship between tonsil defect and lowering of the intelligence quotient, placement of normal tonsils in the "tonsil" group, and of diseased tonsils in the "normal" group would raise the first, and lower the second, thus tending to conceal the difference between the two. On the other hand, the cases where disagreement would occur would naturally be those of slighter defect, in which the intellectual retardation would be less likely to occur, so that the result would probably be merely an increased height at the overlapping portion of the curves, with no change at the ends.
In any case, the two examiners had worked together previously, so that each must have been somewhat familiar with the opinions of the other. They were aware, also, that pronounced tonsillar defect was what we were attempting to detect. However this may be, there must always be some disagreement in diagnosis. When this is allowed for, the results of the investigation may be taken for what they are worth. Contrary to expectation, there seems to be very little difference in intelligence between a group of children whose throats are normal, and one in whom the tonsils are diseased or badly enlarged.
STUDY OF IMPROVEMENT AFTER OPERATION
The complete results of the tests and retests are collected in Table II, where each control case is listed immediately below its respective test case, and where age, height, weight, grip, tapping rate, I.Q., and score in Healy Picture Completion are shown. From these data the more detailed observations have been made. The improvement of each child in the various tests has been computed, and a comparison drawn between the two groups. As we have previously stated, any improvement shown by the test group in excess of that of the control group, may be looked upon as significant.
Let us consider first the improvement of the children in general health, as shown by height and weight. In Tables III and IV we have tabulated the results, in such shape as to permit of comparison. An inspection of these tables will establish the fact that after a six months' interval, the test group shows, in respect to height and weight, a very slight gain over the control group. In weight, the average of the amounts by which the test group gains exceed the control group gains is 1·37 lbs., and in height, only ·16 inches. The medians of these amounts are 1·2 lbs. and ·2 inches respectively. Comparing the improvements for the two groups, we find that in the case of the weights, the smallest gain (a loss of 1·2 lbs.) occurs in the control group, while the largest gain (10·7 lbs.) is in the test group.
TABLE II. RESULTS OF TESTS
Blank spaces indicate where tests were omitted for one reason or another