Five dry battery cells would very readily heat a short piece of fine wire to a sufficiently high temperature to explode the mixture, but it is impossible to alternately heat and cool a wire twelve times a second. It is too slow an operation. The only other method known at present is to imitate the lightning and force an electric current against the resistance of the air with sufficient power to produce the required heat. This, however, requires an extremely high voltage--at least 5000 volts, and our battery of five cells has not more than seven and a half volts of pressure. The interesting question then is, how does the spark coil enable us to raise the voltage from 7 to 5000. To help toward an understanding of the matter I took seven small wire nails which I found in the boat--they were sixpenny finishing nails. I then took two or three yards of No. 24 insulated magnet wire, such as is used upon electric bells, etc. I use it more often than any other wire, and always have some about the boat. I fastened one end of this wire to one of the binding posts of a dry cell (Fig. 113), _a_, and attached branches _c_ and _d_ to it. The other end, _b_, was left free to act as a switch for closing the circuit by touching it to the remaining binding post. [Illustration: Fig. 113] [Illustration: Fig. 114] [Illustration: Fig. 115] One boy then touched the bare ends _c_ and _d_ to the tip of his tongue, while I touched repeatedly the binding post with _b_. There was, of course, no sensation. We now wound a portion of the wire upon the bundle of nails, laying on about fifty turns. (See Fig. 114.) The tongue was now placed at _T_ and _b_ was touched a few times to the free binding post. A very decided shock was felt, not while the end of the wire was resting upon _b_, but at the instant of touching and again at breaking the connection. The shock was noticeably stronger at the instant of breaking than of making the connection. There was also a spark formed when the connection was broken, which did not appear before the coil was made. We next wound on more of the wire--about fifty more turns (Fig. 115). When now connections were made and broken at _b_ the tongue at _T_ felt a much more decided shock, and a larger spark occurred at _b_ when the circuit was broken. Both the tongue and the spark indicate that the voltage is creeping up very rapidly in this series of experiments. We next connected two cells in series, then three, four, and finally five cells in place of the one. The spark grew larger and "fatter," as the boatmen say, with each addition of a cell. It was not pleasant to use the tongue in the experiment after the number of cells exceeded two. I removed the branch _d_ from the wire _b_ and connected it to the binding post, as shown in Fig. 116. I then removed the crystal from my watch and poured into it a little gasolene. I rubbed the ends of _b_ and _d_ together over this, and when they separated the spark which was produced would not light the