They suggest also that we may not only agree with Dr. Rice in his contention that more than fifteen minutes daily spelling drill is time thrown away, but may go farther than he felt warranted in going, and dispense with the spelling drill altogether without prejudice to the educational interests of the pupils. May it not, in other words, be advisable to omit the specific spelling period from the elementary school programme as it has been omitted from the high school and college roster?

An experimental and statistical investigation of this question is presented in the next section.

PART III

PEDAGOGICAL METHOD AND RESULTS

ON purely theoretical, ethico-psychological grounds, Herbartians have maintained the necessity for the correlation of studies. The degree of articulation proposed varies from a "coördination" of several groups of studies to attempts at "concentration" about a central core of material. In their plans of instruction the "formal" branches, as distinguished from the "content studies, are treated as but means to ends, not as ends in themselves. Language (including reading, writing, and spelling) is regarded merely as the vehicle of thought, and is made to follow the development of the latter. Facility in the use of the conventional symbols for the reception and expression of thought is consequently considered to be readily acquired by instruction incidental to the main course.

While not professedly an Herbartian, nor consciously working out his theories upon Herbartian lines, Parker carried the Herbartian principle of concentration to a most radical extreme. He characterizes what he calls "the prevailing hypothesis," as follows: "Pupils must be trained to make the forms of thought, expression, forms of language, art and number with no immediate relation to the thought the forms express. The purpose of this form making is the use of the forms in the future when needed for the expression of thought." This he contrasts with a second hypothesis: "The technical skill necessary for the adequate expression of thought in all modes of expression may be thoroughly acquired under the immediate impulses of intrinsic thought, or the thought evolved in the study of the central subjects and their auxiliaries." The first hypothesis finds expression in such specific statements as: "Constant drills in

oral and written spelling, covering a period of eight years, are absolutely necessary "; whereas the second hypothesis involves the proposition (among others) that "writing, including spelling, punctuation, and capitalization, may be gradually and surely mastered by successive attempts to express thought." The second hypothesis is the one which, upon almost purely theoretical grounds, is strongly advocated. The truth of this hypothesis and the possibility of reliance upon it under actual conditions of classroom instruction have been put to the test of practice in the Cook County Normal School, with, it is claimed, considerable success.

But not only have the coördination and concentration of studies (and the more or less incidental teaching of the formal studies) been advocated, practiced, and defended on such theoretical grounds, but very practical considerations have forced the problem upon the attention of educators not otherwise interested in the Herbartian school of thought. The exigencies of the teacher of the elementary school, who must deal with a curriculum that is ever being enriched by the addition of some new subject matter of instruction, have made the question of economy of time one of great practical importance. Incidental teaching of reading, writing, spelling, etc., has been invoked, in some instances, to meet this demand. But while theoretical (Herbartian) considerations demand the closest possible correlation of "content" and "formal" studies, and practical necessities require the strictest economy of time, the question whether these purposes may be effected by the proposed incidental teaching of the formal branches without prejudice to their interests is one which can be satisfactorily determined only by recourse to actual investigation.

The results of the experimental investigation of spelling errors and the psychological facts and theories bearing upon the acquirement of spelling habits, which were given in the preceding section, suggest the comparative meagreness of the

1 Parker, Francis W., Talks on Pedagogics, pp. 288-290. New York, Kellogg, 1894.

contribution of the specific spelling drill to the final result, and warrant an experiment which might otherwise have seemed a too dangerous tampering with the educational progress of the pupils who were to act as subjects in the tests.

It was decided to abandon the use of the spelling book and home lessons in the subject, to omit also the period from the school programme which had been devoted to its study and recitation, and to investigate the effect that the abstraction of these influences might produce upon the spelling of the pupils of the several school grades. Several methods of measuring results were devised which will be herein described and statistically reported upon. The one most extensively employed was an adaptation of the "composition test" which Dr. Rice employed in his investigations. It consisted of a record of the percentage of accuracy in the spelling of various kinds of exercises that were written in the course of the school years closing June 30, 1898, 1899, and 1900.

Composition Tests. The method pursued was to mark the misspelled words and calculate the percentage of correctness upon the entire number of words written, e.g., a composition paper of three hundred words, of which three were misspelled, would have 99% correctly written. When the same word was misspelled more than once in the same paper, it was counted as an error each time. Dr. Rice allowed such recurring misspelling of the same word to count but once in the total of errors of an individual paper. This would tend, other things equal, to make the results of his tables slightly higher than those herein recorded. The record for a class or class average for any one test was found by taking the median value instead of the arithmetical mean, and the records for separate departments (grammar grades and primary grades) and for the entire school have been obtained also by this method. The use of the median not only furnishes a better representative value than the arithmetical average, but also very materially reduces the labor of computation. The median values of the number of words written in the successive exercises of each grade, department,

and school were recorded, as was also the total number of pupils represented in each result. The results of all the tests of the year 1897-1898 are given in Tables XV-XVIII, inclusive (Appendix, pp. 81-85).

Table XV shows that eight tests were given in the course of the year to pupils from the third b grade to the eighth grade (inclusive). The first, fourth, and eighth tests were exercises in composition; the third and fifth were papers written in answer to questions in geography; the second, sixth, and seventh, similar papers of which language (prose and poetical selections, technical grammar, etc.), science lessons, and history were respectively the subjects. The composition papers of 43 third b grade pupils, e.g., in June, 1897, averaged 81 words, of which 92.4% were spelled correctly; the length of the composition and the spelling percentages increased up to the eighth grade, where 69 pupils wrote on an average 224 words and had 99.5% correctly spelled. Column 9 of the table shows the average number of words written and the average percentage of correctness for the entire eight tests. The final column (“ Ave. Var.") shows the average variation of the spelling percentages of the successive tests from the yearly average. Totals and averages are given for the grammar grades (fifth to eighth years inclusive), for the primary grades (first to fourth years inclusive, the first and second years not being tested), and for the school as a whole (third to eighth years inclusive). The latter, for example, show that 503 pupils taking part in the first test wrote on the average 171 words, with a spelling percentage of 97.1%; that 526 pupils took part in the eighth test, writing 168 words and spelling 97.6% correctly; and that the yearly averages, in which 4779 papers are represented, are 278 words and 97.8%.

In Tables XVI and XVII (Appendix, pp. 82, 83) the same plan of arrangement is followed to show the results for boys and for girls respectively. Table XVIII (Appendix, pp. 84, 85) shows, (a) for boys, (b) for girls, and (c) for boys and girls, results of ten similar tests upon third- and fourth-year pupils