very little. The material accumulated would undoubtedly so systematize and extend our knowledge of these important substances, that we should soon be able to determine in advance all the properties of a proposed alloy, and even to ascertain by calculation what alloys could or could not be formed. The extent of research here suggested may be realized when we remember that, out of the sixty-five elements now known, not one has been thoroughly described, or described with even a moderate approximation to thoroughness. These investigations upon the elements would be for chemistry and physics what the preparation of star-maps and planetary tables is for astronomy, or the dissection of the human frame for medicine. They would certainly furnish a foundation for exact physical science such at at present is scarcely even begun.
After the examination of the elements would come the consideration of compounds. These should be taken up, series by series, in some regular order, and at least every typical body carefully described. Thus, step by step, the lines of assault would be drawn around the besieged problems, until at length the citadel would yield, exactness would replace the present chaos, and definite laws would stand where now are speculations. Could any branch of applied science fail to reap a benefit from this result? Would not every industry in any way dependent upon either chemistry or physics be helped? Apart from direct applications of science to the arts, the mere substitution of accuracy for inaccuracy in questions of scientific principle ought greatly to facilitate technological investigations, put new weapons into the hands of the artisan, and so add immensely to the resources of civilized life.
The investigations here indicated are not by any means the only researches proper for an endowed laboratory. They are merely types, to illustrate the general character of work which such an institution should do. It is true that, although individuals cannot deal with these greater problems in their entirety, individuals may, working separately and disconnectedly, contribute much toward their solution. But, unfortunately, researches of this kind are among the most difficult and arduous. They savor much of hard routine and yield no quick return of glory to the investigator, who, already familiar with monotony in his ordinary duties, naturally prefers to undertake labors producing with less effort a more immediate reputation. The discovery of new compounds is less troublesome, and brings speedier celebrity; hence the more solid work of establishing accurate numerical data is very little done. When done, it is done piecemeal. Garden flowers are so much easier to raise than oaks.
Now, assuming that a laboratory for research ought to be established, let us consider some of the leading questions as to its arrangement and organization. First, with regard to the building. This need not be very expensive, since architectural experiments have no necessary
