are, of course, matters to be judged by the teacher in each instance. It may seem advisable to furnish a brief syllabus, giving, under each heading into which the general subject is divided, a synopsis of points to be investigated or questions to be answered; or, lectures may be given upon the general subject, in order to present more systematically its elementary relations and preserve its continuity, leaving only certain sections to be more fully investigated through the research of the student.

For the purpose of aiding the student in his work, references may be given to books or periodicals to a limited extent, but instruction should also be given him as to the best methods of looking things up for himself. Practice in the use of the various indices, and learning where and how to find what he needs, should be considered of more importance than the number of facts gleaned from the reading. Advice and help should always be freely given the student, that he may work to good purpose and not be obliged to grope in the dark; but the aid should be so rendered as to require the student to do the work for himself.

It would be useless to enter into a discussion concerning the details of arrangement of a seminar course or the variations to which the method is subject. The conduct of the work must be varied in each instance according to the capabilities of the teacher, the nature of the subject, and the number and training of the students in the class.

The object in presenting this paper to the Society is simply to invite attention to the desirability of intro

ducing into engineering courses some means for making the student better acquainted with the range and use of engineering literature and to suggest that the adaptation of the seminar method to some of the more advanced studies of the course might be of very great advantage as conducing to this end.

In many engineering schools efforts are made to interest the students in the study of engineering literature, various methods being used for that purpose. References are given in lectures to matters of general or historic interest; the reading of current periodicals is required or papers are prepared by the students upon assigned topics. It is undoubtedly true, however, that a large proportion of those who graduate from engineering courses have very little idea of any professional literature beyond the text-books they have used, and perhaps a magazine or newspaper they have become accustomed to read. Looking up references given in lectures and the reading of current periodicals are both of value in stimulating a taste for general reading, but they have not the same office as the systematic study of literature relating to a single subject. The preparation of papers for the engineering societies which exist in most colleges, and the thesis work commonly required for graduation, both ordinarily involve some research upon the part of the student and tend to stimulate a taste for independent investigation; but, as such work is not commonly under the immediate direction of the teacher, the student has not the benefit of that careful guidance which may be given in a definitely projected course, and is not held to the same accountability for the character of his work.

In considering the introduction of seminar methods. it is at once clear that they can have but very limited application in an undergraduate engineering course. Evidently a method such as that outlined above would not be applicable to those fundamental subjects which form the base of an engineering education and which constitute the major portion of the work of an engineering course. These require logical and orderly development as well as the use of methods by which they may be given briefly and concisely.

The seminar methods should be limited to a few of the subjects near the end of the course, preferably those which may properly be called "professional subjects."

These methods also are not well suited to a very brief course. It requires more time for the student to obtain the same knowledge of the subject treated than if it be given by lectures or from a single text, while in order to accomplish good results sufficient time should be available to go quite fully into the subjects considered. They may therefore be applied with advantage to certain lines of professional work into which the student may be permitted to go more extensively than is possible in others. It is still more an advantage if the work so treated may be to a certain extent optional, and all students not necessarily held to the same work.

It matters very little what the particular line of work may be, provided the student be required to go somewhat fully into an examination of the fragmentary literature of the subject chosen, and to put his results into a shape available for use. Having investigated one subject he will know better how to proceed when under the necessity of investigating another.

The object, as already stated, should not be primarily to communicate knowledge of a special subject, although the information obtained must be of value, but rather to give an idea of the extent and nature of the literature of engineering, and to cultivate, so far as may be, the ability to use it.

DISCUSSION.

PROFESSOR DE VOLSON WOOD wrote that he thought the suggestions in this paper to be worthy of serious consideration. The success of the method depends upon "environment." He had been able to do but little, very little, with it with large classes, except in the preparation of theses. There are two courses in the Institute which may be called "Seminar." One is in a supplementary term of four weeks in which certain classes do shop work, make tests of all kinds of machines, take and reduce notes and the like. The class is divided into many sections that there may be only a few students in a section; and temporary instructors, most of whom are employed for the four weeks, take charge of special subjects. One may have charge of refrigeration, another hot-air engines, and so on. During this time there are no recitations, but at its close there is an examination. This is considered an exceedingly valuable part of the course. It will be seen that the plan is similar to that described in the paper by Professor Marvin in Civil Engineering. Our other seminar is the preparation of a thesis for which at least eight weeks are given at the end of the senior year. The student is put upon his own resources, but upon any new or difficult point

is free to consult with the professors. When completed it is reviewed by two or more professors, and notes made in regard to the exhibition of knowledge, clearness of descriptive matter and neatness of execution, and if not satisfactory it must be revised. When satisfactory it is typewritten and a copy left with the Librarian. It is quite possible that a seminar following more closely the method of Professor Spalding would be more profitable in many cases than the thesis as described above, but in his own case the method used by him had, he thought, given a better average result up to the present time than any other plan would be likely to.

PROFESSOR F. E. TURNEAURE wrote that he considered the seminary in engineering education one of the most valuable aids at our command. Education means the ability to investigate and to draw correct conclusions, and this implies, first, the ability to study and reason, and second, practice in getting and using information and therefore, as well, a knowledge of the sources of information. Lectures and text-books are necessary in teaching the tools for use in investigation and they may furnish a considerable amount of valuable information, but after a knowledge of these tools is acquired, the seminary method, whether called by this name or not, is essential to the acquirement of power and independence. Engineering graduates are often needlessly handicapped simply from lack of knowledge of the sources of information, and this lack may reduce the efficiency of their work for several years. This certainly can be corrected by even a moderate amount of seminary work. The use of as