fringes. We could not, in the time, build another apparatus of timber which had not been dried all winter, nor was it thought well to construct another apparatus closely resembling the first. While planning a new apparatus, we made a couple of experiments to show, what was well enough known, that difference of magnetic attraction on the iron parts of our apparatus could not disturb our observations. We suspended two massive pieces of iron at the ends of one arm, so that one should be in the lines of magnetic force of the earth's field, and the other transverse to them, these relations being reversed on reversing the position of the apparatus. But observations with this load of iron gave the same result as before. Next we placed an analytical balance on one arm, with which to weigh a bur of iron at the extremity of that arm. It was so placed that at one azimuth the bar was nearly in the lines of force, and at another was transverse to them. If there were a difference of half a milligram in twelve hundred grams, it would have been detected, but no such difference existed. We found by trial how much a weight of a hundred grams displaced our fringes, and so learned, as was known before, that the influence of the earth's magnetism could not be a disturbing factor.
The Rumford Committee of the American Academy of Arts and Sciences having made a grant in aid of this experiment, we carried out our original plan of making a steel structure so rigid as to permit easy and satisfactory observation. In this new apparatus all the optical parts are carried by a steel frame built of plate and angle-iron, somewhat like a bridge girder. A cubical steel box, fourteen inches on each edge, constitutes the centre of the structure, which is in the form of a cross. To each of the four sides of this cube are firmly attached arms, each about six feet and a half in length. Each arm is made of steel plates, three eighths of an inch thick, eighteen inches wide at one end, and six inches wide at the other, standing on edge, and kept fourteen inches apart by suitable plates, angle-irons, and other braces; thus are formed hollow beams of great rigidity, especially in a vertical direction. This framework is shown in Plate 1, from which it is seen that the structure is in effect two rigid beams, each fourteen feet long, crossing at right angles, and symmetrical as regards strength and rigidity.
On two ends of the cross, S and T, Figure 1, are two upright cast-iron frames, fastened by bolts, each of which carries four mirrors, marked 2, 4, 6, and 8. Against the corners of each of these frames rest four pine rods, three quarters of an inch in diameter and fourteen feet long. Each rod is supported throughout its length by a brass tube an inch in
