whole experiment water is being lost by evaporation. When we consider all these corrections that must be carefully made as well as the fact that to accurately read the height of the mercury in the thermometer it would probably be necessary to look at it with a telescope, the difficulties in this simple experiment and the temptation to slight something are very apparent, and yet this is what we expect a freshman to do in the time of about two hours in the laboratory, and at the same time we expect his result to have an accuracy considerably better than one part in a hundred.
The second and in many cases far more important function of the laboratory is to serve as a place for the performance of accurate research, that is, the investigation and discovery of new phenomena. In order to take part in this inspiring occupation it is obvious that the student must have acquired a considerable amount of proficiency and have already made measurements of a great variety involving a high degree of precision. It is often supposed that scientific discoveries are attended with a large amount of luck, or that they are the result of a sudden inspiration which may come to anybody. Such is far from being the case. Professors of physics are frequently the recipients of visits from persons who in their enthusiasm feel that they have made an important discovery, which in many cases has been thrown off as a sort of by-product in some other vocation. Not many years ago I received a visit from a young man who had traveled over two hundred miles to present to me the results of a theory which he had elaborated to account for the motion of rotation of the planets on their axes. After I had inquired whether he had made himself familiar with the writings of the great masters in celestial mechanics, and had explained to him the impossibility of his theory, I asked him this question, "Do you realize, my dear sir, that if your theory were correct, it would upset the consequences of all the astronomical observations that have been made during the last two hundred years?" The young man went away sadder but wiser and I did not hear from him again. As a matter of fact discoveries are seldom made by persons not possessing the training that I have described, and in ninety-nine cases out of a hundred the element of chance is reduced to the smallest possible dimensions.
I have already stated that the provision of great physical laboratories in connection with instruction is extremely modern. In England the pioneer work in systematized instruction was done in Oxford and London about 1867. The Clarendon Laboratory at Oxford was built from 1868 to 1872, while the Cavendish Laboratory in Cambridge was, as has been stated, not opened until 1874. In this country the first systematic laboratory course in physics was organized about 1870 by Professor E. C. Pickering, at the Massachusetts Institute of Technology, which illustrates one of my points that I have already made, for Professor
