This piece of iron wire offered resistance to the flow of the electric current. It offered resistance to the motion of the dynamo. This offered resistance to the steam-engine which drives the dynamo. This caused the governor of the engine to open and pass more steam from the boiler. This reduced the pressure at the steam gauge. This caused the fireman to shovel more coal into the furnace. The heat of the burning coal melts the wire, but it does it only after several changes. First, it is converted into mechanical energy in the steam-engine with great loss--about nine tenths being lost. Second, it is converted into electrical energy by the dynamo, with some loss, and, third, it is conducted to the iron wire and converted back to heat with still further loss. It is evident that the most economical way to heat the wire would be to take it to the furnace. Yet all electric cooking is done by sending electric current through wires embedded in the walls of the cooking utensils, and it is the most wasteful method of using the energy stored in coal that has yet been devised. [Illustration: Fig. 68] That merely connecting the binding posts _a_ and _b_ (Fig. 67) by a small piece of wire should throw a load upon the dynamo miles away; should offer resistance to its motion, and make it require 1.18 horse-power more of energy to keep up its speed of revolution, is, indeed, uncanny. I will attempt to make it seem more real. At one end of the lecture table I have a rotary pump _P_ (Fig. 68). The end of the rubber tube _a_, which leads to the pump is lying upon the table outside of the tank of water, _T_. While things are in this condition I move the crank which operates the pump with perfect ease. Now while still turning the crank I pick up the tube _a_ and drop its free end into the water tank. I cannot now conceal the fact, even if I were disposed to do so, that I must work hard to keep the pump going. The pump itself tells you by its laboured sound that it is working hard, and the stream of water which issues from the pipe _b_ tells how much work I am performing. The pump is discharging five and a half pints of water per second, that is 5.5 pounds, and it raises this water 10 feet. Hence I am doing 55 foot pounds of work per second, which requires one tenth of a horse-power. Here is a lad who consents to try the experiment for us. He turns the crank easily while I am holding the tube _a_ out of the water, but when I lower it into the water he finds the resistance so great that, tug however much he may, he is unable to keep the pump going.