Provide yourself with the model of a house made of tin, four inches in breadth, six long, and about five in height. Let there be a chimney placed in the roof equidistant from both ends, and let a glass tube pass through it, the upper extremity of which must reach a little above the chimney, and the lower one come within an inch of the floor of the house.—Let a small wire pass through the bore of the glass tube, the upper end of which must extend a small distance above the orifice of the tube, having its extremity, which must be pointed, furnished with a screw, on which a metallic ball is to be fastened. The other end must likewise have a ball fixed upon it.—The instrument being thus prepared, fill the house with cotton, and sprinkle a little powdered rosin on that part of it, which is immediately between the lower knob of the wire, and the floor of the house. Then connect the lower part of the instrument with the outside coating of a pretty large jar.—From the prime conductor, in order to represent the clouds, suspend a small scale beam, having two balls of metal or wood coated with tin foil, in the place of the scale dishes, nicely balanced. The knob of the jar being connected with the prime conductor; bring the ball on the wire extending through the glass tube, under one of the balls representing a cloud.—Now charge the jar. The cloud will be attracted by the ball on the wire—the electricity of the cloud will be discharged—and if the experiment succeeds, the contents of the house will be set on fire. _The effects of lightning, when a house is furnished with a pointed conductor._ Repeat the above experiment with this variation: unscrew the ball from the upper extremity of the wire of the house, so that it may remain pointed. Place the house under the cloud as in the former experiment.—You will now find it impossible to charge the jar: or if you charge the jar before the house is placed under the cloud; the cloud, instead of being attracted by it, will be repelled, and the jar will be discharged without any explosion, and without firing the cotton. These two experiments evince that _pointed_ conductors are more proper to secure houses from the effects of lightning that those terminating with a ball or knob, and that if the pointed conductors fairly act on the cloud the security is complete. CHAP. XVI. _Of animal electricity._