connection with the purposes in view. It should be plain and substantial outwardly, sufficiently spacious within, accessible to much sunlight and away from the heavier jar of traffic. The suburbs of a great city would be perhaps the most advantageous position for it to occupy. The most serious considerations, however, would concern its interior arrangements. It should, of course, contain a sufficient number of rooms for the accommodation of different branches of research; for example, a photometric room, another for gas analysis, a third for electric measurement, a fourth for calorimetric work, and so on. In the basement, connecting with the solid earth, might be placed a number of heavy stone piers for the support of very delicate instruments. One important item of apparatus would be a steam-engine. This, together with the chemical furnaces and a small machine-shop, might be provided for in a cheap out-building apart from the main structure. Steam, gas, and water, should of course be available in all parts of the laboratory.
But although expense could be avoided in the building, it ought not to be dodged in the purchase of instruments. These would necessarily be of the most costly character. Mere models, such as are commonly used for class instruction, would not suffice. Every instrument of precision used in the laboratory should be a standard of its kind, the best which could be made; otherwise the work of the institution might fall short of the high character intended for it. So also with the chemicals: only the best should be tolerated. As for a library, fitted for scientific reference-work, the cost would depend much upon locality. In a country town, away from other collections of books, the expense would be considerable; but near a city provided with libraries the outlay need not be very great. Still, some money would have to be expended in this very important direction.
Next as regards the working-staff. Since the researches to be undertaken are mainly those which involve the cooperation of specialists, we must start with a sufficiently large and varied body of men. At the head of the institution there ought to be a man of thorough training, proved ability, broad general ideas, and great executive capacity. He should guide and systematize all the work of the laboratory, and to him, as to the director of an observatory, the others should be subordinate. Under him should be at least the following corps of principals: one chemist, one electrician, a specialist in heat, another in optics, a mathematician, and an expert mechanic. Upon the last-named individual would devolve the duty of constructing, altering, or repairing apparatus. To this main staff might be added assistants, as many as the means of the laboratory would allow. Some of these minor positions could perhaps be filled by means of fellowships, analogous to those recently established by the Johns Hopkins University. It might be feasible also to admit private investigators and post-graduate students to the advantages for research afforded
