The words written by the, higher grade pupils being longer require a greater number of coördinated movements for their execution. The increase in rate of movement as measured by number of words written is therefore relatively greater than that represented in the tables or curves. This increase in rate

777 boys: per cent spelled correctly, 89.6 to 97.3; average number words, 79 to 245.

780 girls: per cent spelled correctly, 94.6 to 98.4; average number words, 81 to 212.

is a complex phenomenon and may be regarded as representative, not only of gain in manual dexterity, but also of growth in available vocabulary and of development in rapidity of ' ideation.

Figs. 3, 4, and 5 disclose a certain degree of correspondence between rate of movement and correctness of spelling, i.e., as

the rate of movement changes there is, in general, a corresponding fluctuation in the spelling curve, though growth in rate of movement shows, on the whole, a somewhat steadier gain. This would seem to show that the more rapid workers are also the best spellers. This result seems to hold as between boys (Fig. 4) or girls (Fig. 5) or mixed classes (Fig. 3), but the converse is true when boys and girls are compared with each other. If this rate is conditioned by facility in thinking and by richness of vocabulary as well as by the physiological factor of the innervation of the musculature of writing, the correspondence alluded to would seem to indicate a certain correlation between general mental efficiency and correctness of spelling, for rapidity of ideation and acquirement of large vocabulary are, other things equal, criteria of intellectual ability. Dr. Rice arrives at the same conclusion as the result of a direct investigation of the question. He says, "The lesson to be learned from Table No. 2 [in which his statistics on this point are presented] is, that an unusually high or low class average may now and then be accounted for by an exceptionally bright or dull set of pupils. Occasionally, therefore, the teacher may be allowed to plead 'dull pupils' as an excuse for poor results." 1

These statements could be confirmed from the records of some of the classes whose results are presented in the various tables. Some of the irregularities of the spelling curves are no doubt to be attributed, in part at least, to the unusual segregation of exceptional pupils in a class. Of course the relation posited is only a general one; the exceptions are numerous and important, and good spelling should not be made a criterion of mental ability in individual cases.

That the boys wrote a greater number of words than the girls, though the latter excelled in the spelling, would seem either to negative the preceding conclusions or to point to the general mental superiority of the girls in spite of their slower rate of movement. Neither of these alternatives, however, must be accepted. Other important results of the tests must be considered 1The Futility of the Spelling Grind," Forum, June, 1897, p. 411.

in conjunction with those already discussed. (Reference to Tables III, V, and VII will show that the boys wrote more nonsense words than did the girls and very much more illegibly ;) their ratios of illegibility (measured by the percentage of illegible words written) were 9:1, 14:1, and 6:1 for the three tests respectively. That is, the boys followed implicitly the directions of the experimenter "to write as many words as they could and whatever they happened to think of." They were more interested in the quantity than the quality of their productions and so sacrificed spelling and legibility to the end of producing a large number of words, On the other hand the girls missed this result to a certain extent by their conscientious attention to the details of execution.) This is manifested in some papers by the careful erasure of errors and substitution of corrections, and also by the records of illegible and nonsense words above referred to. Some confirmation of these conclusions will be found in the next part, where the spelling errors are subjected to analysis. The most effective distribution of attention between a final end or purpose and the successive secondary steps which mediate its accomplishment is a problem of very nice discrimination, and its solution a great success. The differences between the girls' and boys' papers seem to be indicative of interesting sex tendencies in this regard.

To summarize briefly the results of the study of the experimental word lists, we may say:

(a) That pupils of the elementary schools increase regularly from grade to grade in the quantity and quality of their spontaneously written words, and in accuracy of spelling.

(b) That the average results established by this method are constant within small limits of variation and may be utilized as normals with which to compare the work of individuals or of classes, under varying pedagogical conditions. (Tables VI and VII have a special significance in this connection, discussion of which has been reserved for a later place.)

(e) That boys show more rapid rate of movement (a complex of motor, ideational, and verbal elements) than girls, but the

latter excel markedly in both legibility and correctness of spelling. The pedagogical corollary may, perhaps, be drawn here that boys need training in minutiæ of execution, attention to details; girls, on the other hand, requiring stimulation to the accomplishment of larger purposes.

(d) That rate of movement and spelling accuracy tend to vary together, and both are functions of general mental capacity.

(e) The fifteen-minute word list (abridged for convenience to a ten- or five-minute list) has a positive value as one of a standard series of tests of the mental and physical characteristics of classes of individuals. It furnishes, in permanent and readily consultable form, data concerning rate of movement, fatigue, legibility and other characteristics of writing, vocabulary, and proficiency in spelling.

PART II .

ANALYSIS OF ERRORS

CAREFUL examination of any list of misspelled words, but especially of lists such as those procured under the conditions of the fifteen-minute tests, discloses certain types or classes of frequently recurring errors. The characteristics of these classes are such as to evidence the necessity and value of regarding written spelling as being subject to the same conditions, anatomical, physiological, and psychological, that are usually recognized in the analysis of articulate speech and allied functions.

Since Broca's epoch-making paper (1861) there has been the greatest activity in the study of these functions, especially from the clinical and pathological side, though related psychological considerations have not been neglected. The knowledge of facts, and the development of theories concerning the factors involved in the reception of the sensory elements and in the exteriorization of thought in various expressive forms, which are at hand as the result of this activity, may be consulted with advantage for assistance in the analysis of material furnished by orthographical errors. As a preliminary, therefore, to the detailed qualitative treatment of these errors, and introductory to the scheme of classification which it has been found advisable to use in order to present a quantitative statement of results, it will be instructive to consider briefly the general subject of aphasia as now understood.

The term was first applied to the inability to express thought in words. "But," to quote from a recent work by Dr. Joseph Collins,1 "it was not long before it became apparent that a word was needed not only to connote restrictedly an inability to 1 Collins, Joseph, The Faculty of Speech, pp. 2, 3. New York, Macmillan, 1898.