Figures have been prepared giving the exact relation of the size of rubber to the number of turns, although such statistics are elastic. The problem may be worked out with your own motor. Differences of temperature will be quickly noted. The rules prepared by V. E. Johnson, M.A., an English authority on aviation, are as follows: The motive power is doubled by increasing the number of rubber strands one-half; by doubling the number of strands, the motive power is increased more than two times; and the tripling of the strands increases the motive power seven times. As regards the number of turns the same authority states that the doubling of the number of strands diminishes the number of turns by one-third to one-half. [Illustration: A Metal Skid] It is also found that each strand will have doubled knots of 310 turns; four strands, 440; sixteen strands at 200; and eight at 210. This is working with strands one-sixteenth of an inch square. As a rule, rubber should not be turned after the second row of knots is formed. And by the way, you will find that the rubber, after being tightly twisted, tends to stick together, and should be carefully separated after a flight so that the air can reach all surfaces. According to the experiments made by Mr. Johnson, one pound of rubber may be made to store up 320 foot pounds of energy, while one pound of steel, in the form of springs, will only store up 65 pounds. In the early model aeroplanes much valuable energy was lost through friction. There has been a marked improvement in the construction of the propellers, axles, and bearings. Friction has been reduced to practically nothing. It is possible, of course, to drive a propeller with the shaft turning in a hole drilled at random through a stick, with a glass bead for a washer. It is very important, however, that the axle should turn exactly at right angles, and to hold it in position requires careful adjustment. To meet the demands of model aeroplane builders, several shaft mechanisms have been prepared, even to a very complete arrangement of miniature ball bearings. The model builder who cannot avail himself of these parts can, nevertheless, imitate their action with reasonable fidelity. The axle attached to the propeller should be heavy enough to resist bending in ordinary wear and tear. A bicycle spoke is just the thing. When you have decided upon this axle, procure a piece of metal tubing in which the axle will turn freely, without binding or rattling about. The tubing should then be passed through the frame supporting the propeller exactly at right angles, and extend out at either side about half an inch. To fasten it securely in position, glue and if necessary drive small wedges,—a match or toothpick,—about it.