The sling shot device for throwing the glider is made of heavy spring wire, and will require a strong metal vice to bend it in. A forked stick can be used, or one can be cut out with a turning saw. Fig. 218 is quite similar to Fig. 217 in weight, but a square spine ¼” × ⅜” × 18” with planes ⅛” or 1/16” × 2” × 6”. 2” in widest part and the other ⅛” or 1/16” × 4” × 12”. 4” in widest part. The force used to drive this glider is given thru a springy stick of some tough wood, as oak or hickory. The stick should be quite stiff so as to resist more pressure before its release. A little block on the under side of the spine might have a little hole in the back to receive a small nail in the end of the bow stick used to throw the glider, this will prevent the bow stick from slipping off in the throwing process. One other glider should be mentioned, and that is one with sheet metal planes, Fig. 219. This has advantages and disadvantages. When it strikes hard against some object, the metal is liable to bend, also if it is thrown violently, and should strike someone on the face or hands it might cut. The metal surfaces can be bent into any shape. All corners should be rounded. Some of these gliders can be thrown long distances if properly adjusted. After working awhile with gliders, we can try model aeroplanes that are suited for the instalment of motors later. Everything must be made as light and strong as possible. All kinds of ribbed surfaces, keels, and light wire braced frameworks, are utilized. Everything that would be used in a self-propelling model, except the propeller and motor. Make the planes movable so they may be balanced as to pressure, by moving them back and forth, flex more, flex less, tilt more, tilt less, until you get a good glider out of it, then attach your motor and propeller. Some may think best to put motor and propeller on, and do all the experimenting, but it takes time to make a good propeller, and the fewer jams it has the better, so it is better to do some experimenting with the model before the propeller is added. The motor will make practically no difference in the balance, so there would be no advantage of putting it on for experiments in gliding. We are now ready for some attempts on the self-propelling models. CHAPTER XV. MODEL AEROPLANES. =Self-propelling air devices= must be of the lightest possible weight and yet they must have strength. There will be no attempt to give a great variety of model aeroplanes; but a few can be discussed, perhaps three or four. The first is a little practice model that has been of great service to me, both for study and for instruction to others. It is small and will not fly far. I can wind it up and turn it loose in a schoolroom. It hits rather hard sometimes but nothing has happened to it beyond a broken propeller, and the rubber bands that are used to tie the planes to the spine have been broken. [Illustration: FIG. 220.]