For my curvometer, a difference of reading of 40 divisions represents a difference in radius of the curve of 82 inches. This means that if the curvometer were applied to a perfectly flat limb, and the limb bent until the radius were 82 inches, the dial reading would be 40 ten-thousandths. Likewise, for a reading of 50, the corresponding radius is 65.5 inches; for 100, 32.7 inches. The values can be found from the formula *R*= (*A*^{2}+*D*^{2})/2*D*, where *R* is the radius of curvature, *A* half the distance between fixed points and *D* the displacement of the feeler point of the gage, as shown by the dial reading.

A dial gage of the kind described is quite expensive, so I made it serve two purposes, both important in the testing of bows. The first has been discussed. The second is the weighing of a bow with a high degree of accuracy. Figure 2 gives a partial view of the bow weigher with a bow in position for weighing, at about half draw.

Making the bow weigher was suggested to me by seeing Forrest Nagler's. He had cut the frame of his out of boiler plate, and used a dial indicator reading thousandths to measure the deformation of the frame produced by the drawn bow. Being so much lazier than Forrest, I couldn't see myself scrollsawing a shape out of boiler plate; so I decided to let the bow compress a ring of steel or hard brass, and measure the amount of compression with the dial gage. An advantage of the ring is that it can easily be replaced by another to change the weighing range, and it can be adjusted, by adjusting width, to give readings in kilograms or pounds. As I have it, the range is 143 pounds with compression of 130 thousandths, or 1300 divisions. This corresponds to five kilograms or eleven pounds per 100 divisions. It measures accurately to plus or minus 0.1 pound over the entire range.

A plate fixed to the ring serves as a bow rest. At the opposite diameter, a bracket for the dial gage is mounted inside the ring, and another bracket on the outside supports a 30-inch rod, along the length of which there is a peg every two inches. By hooking the string over each peg in turn, data for the force-drawn curve are quickly obtained. The rod with the pegs may be detached from its holding bracket by unscrewing, for greater portability.

In both curvometer and bow weigher, the dial gage is secured to a stout bracket by means of two screws. The gage is easily transferred from one to the other. When it is used as curvometer, the feeler point is fitted with a rounded end, as shown in Fig. 1, to give a point contact and to prevent marring the bow by the sharp edge of the point.