Why did not the man who built our mill two generations ago locate it upon the small stream that flowed near his house? The small stream was more conveniently located for him and it has quite as much fall as he got at the foot of this lake. We sometimes express the fact by saying that the "head of water" or the water pressure was quite as much in one of these cases as the other. One boy said that the stream sometimes gives out. Another one said that it never did have water enough to run that wheel. "Undoubtedly the trouble is with the quantity," said I, "but I want to show you that we cannot maintain the pressure unless there is sufficient quantity back of it." [Illustration: Fig. 123] In Fig. 123, suppose _A_ represents a small, slim tank of water three feet high. The water-wheel _W_, requires one gallon of water a minute pushed along by a three-foot head of water pressure to run it. The supply pipe _S_ is bringing into the tank not more than one quart of water per minute. A gate at _R_ enables us to regulate the flow of water, as we regulate the flow of electricity, by using more or less resistance. Now it is evident that if we close the gate, or partially close it, and allow the tank to fill with water, we may then open the gate and run the wheel for a short time, but the level of the water in the tank soon begins to fall and the pressure grows less and the wheel stops moving. It is just so with all generators of electric current. If we take from them more than they can supply continuously the voltage falls. This is notoriously true of dry cells. Like the water tank represented in Fig. 123, they "run down" if used continuously to furnish, say, one ampere of current, but they may furnish it for a short time, the voltage rapidly falling meanwhile. Then if given a short rest they "pick up" and will again furnish full pressure. The voltage of a dry cell falls somewhat when it is required to give the very small amount of current required to actuate a volt meter, say .015 ampere. Hence, our volt meter will not quite correctly show what the voltage of a single cell would be on open circuit. Notice that, when I put one cell upon this volt meter the needle shows 1.42 volts; but when I put four cells in series upon it the needle indicates six volts, as nearly as we can read it. That is, the voltage of each cell in this case appears to be 1.5. What has increased the voltage of a cell from 1.42 to 1.50? Simply this: when .015 ampere, the amount required by the volt meter, was taken from one cell it reduced its pressure, but when a multiplier with ten times the resistance was added we secured our reading by using only .006 ampere of current, and this did not appreciably reduce the true pressure of the cells.