cial considerations rarely intervene. No precaution to insure accuracy, in fact, need ever be neglected.

In conclusion it may be said that it is certainly true that in the past, investigations of the greatest value to the engineering public have been conducted by students under the auspices of college instruction. It can be confidently predicted that the superior equipment of colleges and the better training of instructors and students will lead to better work in the future than has been done in the past. It is believed that in the future, as in the past, many of the engineering coefficients which are to be determined, and new engineering laws which are to be developed, will have for their foundation the original work done by students in engineering colleges.

DISCUSSION.

PROFESSOR D. S. JACOBUS thought probably the author of the paper had referred to him when he said that his paper had been brought out by a criticism that had been made on the subject at the meeting of the Mechanical Engineers. The point there raised was that it is not wise to put up an apparatus to assist in instructing students and then consider that the results of the students' work are good enough to plot down to determine some physical law. Of course in exceptional cases the student may be entrusted with some original work; after he has worked some four months and has thrown away all his work, he may have come to the point where his work is of value. But when it comes to rushing through a lot of students and trying to deduce some law from the results, the speaker did not think

it could be done. In all the commercial work done by the speaker and his associates they never trusted the obtaining of the data to the students. They have a special corp of assistants who have been employed for years, and they find it is a regular trade to be able to note everything aright and practically never make a mistake. Again, in obtaining this data the ones that are most likely to make mistakes are not the dullest ones, because if a man is thinking too much he may make more mistakes than another man who only thinks just about enough. The speaker gave an instance where an ignorant Irishman was found to be absolutely reliable in counting the strokes of an engine where more intelligent intelligent men would always make mistakes. It don't follow because a man is brilliant he is never going to make a mistake. He considered it a regular trade to take all the data, and he had special assistants who were used to that class of work, and in whom he could place a great deal of confidence, but if you undertake to have a party of students do work which they are not used to he thought it very risky business.

PROFESSOR H. W. SPANGLER wished to say a word about original research by undergraduates. Professor Marx had said in effect that the results are reliable because we have competent observers and because the apparatus is complete. Those who practice engineering in addition to their college work realize that there are few who have the capacity to set up apparatus and carry out a series of experiments. The speaker believed that in the short time in the

latter half of the senior year it is practically impossible to do this sort of work. It can be carried on from year to year, but where are the students? It is not a research by the students, but observation by the students. The students are a set of observers and it is simply out of the question to put any credence upon results based upon experiments of this kind. He had had some little experience and it would be a very unusual case where he would put a corporation representing $200,000 into a line of work as the result of a series of experiments carried out in any educational institution that he knew of, and he believed that any corporation that had a large amount of money to put into an experiment would not undertake to do it on results reached by a class of students. He could not think any one would be willing to put credence in results arrived at in that way. These young men do the best they know, but they don't know how it takes time.

The speaker wished to say a word as to the necessity of liberal study in connection with engineering work. Our engineers go into positions where they meet men of liberal education, men who can talk of other things besides engineering and talk it well, and he thought that many of our institutions are making a great mistake in cutting down the liberal portion of the students' education. He did not mean to say the study of engineering is not liberal in its way, but in a broader sense we are cutting it down so that our men are going to work in a very narrow field and they will get into a small rut in a few years. It seemed to him that the course

of work outlined by Prof. Carpenter left the experimenting to skilled observers, not to engineers. But the business of an engineering institution is not this class of work. It was evident that if under his direction a series of experiments are being made, if he was responsible for that work, he proposed to see that the work was carried out in the right way; these men were to do what he set before them. If he was to be responsible, then he would simply drill those men, he would have them do the work he wanted them to do in his way. He could do that in such a way that they would get the idea that they were doing it, but the fact would be that if he had to give an engineering opinion on that work he would not take the results of others, he would take their observations and satisfy himself first that those observations were right. Beyond that he was responsible and it was his work, not the work of his students. Therefore he would be developing men to become observers, or experimentalists, in a very narrow sense. It seemed to him this course of work was strongly in that direction, and therefore, he thought, unfortunate.

In research work by students he had a very distinct idea as to what he wanted to accomplish. He wished to make all the men alert, quick, sharp, exact, and to develop if possible their engineering sense. That is all. He had no other object in the research work that he carried out than those three things. If he could do that, he had given them a broad enough foundation, he believed, to make good engineers in the future.

PROFESSOR CARPENTER was not sure that he differed particularly from Professor Spangler. He did distinctly believe, however, everything stated in the paper. He agreed with Professor Spangler that research work should not be undertaken until the student had had two years' practice. If, after two years' practice, a man is not a skilled observer, if his opinions are not worth anything, where will we find them? Students are bright as a rule, they are well trained for the work, and the work they do is good, much better than that done by outsiders. The laboratory exercises are arranged with the idea of making the students more intimate with the laws of forces and materials. The under-graduate laboratory, as we conceive, it is not run for the mere purpose of making observations, but to give the students greater familiarity with the machines, giving them an opportunity to handle and study the machines, which they cannot get from text-books, and the result is that after they have had that course they go out into the world very much better prepared to meet the problems of engineering life, very much better equipped.

The speaker had not referred to Professor Jacobus in his paper, and he perfectly agreed with him, and some experience of his own had been exactly in accord with that of Professor Jacobus. He never yet had been able to utilize that work and did not see how it is possible to arrange such work so as to be available. However, he did not refer to that class of work. He might say that in several instances with under-graduate work, for instance in investigating the properties of