In this case, the knob of the jar is charged positively, and the spider, being in its natural state, is attracted by it; the knob then communicates to it some of its electricity, and the spider becoming possessed of the same electricity with the knob, is repelled by it, and immediately runs to the other knob, which communicates with the negative coating, or with the table, where it discharges its electricity and is again attracted by the knob of the jar. This attraction and repulsion continue till the jar is discharged, when the spider finishes its motion and seemingly expires. CHAP. IX. _Of the Leyden phial._ This consists of a glass phial, jar, or bottle, coated on the outside and inside with tin-foil, rendered adhesive by paste or gum water. About two inches of the glass at the top are left without any metallic covering, to prevent a communication between the outside and inside coatings, while the electricity is collecting.—The mouth of the phial or jar is furnished with a cork which receives a wire, ending in several ramifications which touch the inside coating. The upper end of this wire, which should extend a convenient distance above the mouth of the jar, is furnished with a metallic ball. When the phial or jar is to be charged, it may be held in the hand or placed on an uninsulated table, with the knob of the wire touching the prime conductor. The inner surface of the glass now acquires the same electricity with the prime conductor, and the external one acquires a contrary electricity by means of its uninsulated coating. When a phial similar to the one above described is highly charged, a spontaneous discharge will usually take place over the uncoated surface, and seldom through the glass. But if the uncoated surface be left larger than from two to three inches, the phial is more apt to crack and become useless, by the charge passing through the glass. There is not however an absolute certainty that a jar which has once discharged itself over its surface will not, at another time, break by a discharge through the glass. It was long disputed whether the discharge of the Leyden phial resided in the coating or in the electric. The following experiment clearly decides, that its residence is in the electric. Upon an uninsulated plate of metal, lay a plate of glass considerably larger, so that there may be a rim of three or four inches projecting beyond the metal. Upon the glass lay another piece of metal, of the same size with the first, and so as precisely to cover it. Let this instrument be charged, by connecting the upper metallic plate with the prime conductor. Then separate the metallic plates from the glass; and upon examination the glass will be found to possess the contrary electricities on its opposite sides; that side which during the electrization communicated with the prime conductor will have a like electricity with it, and the other the contrary.