Roberts looked up a little surprisedly, but a few words from Fields explained the situation, and the newcomer turned to Roger quite affably. "Certainly, my boy, it's very simple," he said. "You see, all work on rivers, whether for the purpose of irrigation, flood control, or navigation, is dependent on the amount of water that flows through that river channel every year. A week of wet weather makes a vast difference to the amount of water the river is carrying, and a dry spell cuts it off." "But don't springs and things keep the water about even?" queried the boy. "No, indeed," answered his informant emphatically. "Why, the Tennessee River, which I worked on once, for three months never flowed more than 20,000 cubic feet per second, yet that same year, for fifteen days in the spring, it tore down with over 360,000 feet a second. In other words, in the spring it was as big as eighteen rivers its usual size." [Illustration: MEASURING STREAM FLOW. Trolley line one mile long, over an Eastern river. Instruments pulled up, ready for return to the bank. _Photograph by U.S.G.S._] "Are all rivers like that?" "Most of them. You see, suppose in the middle of summer a river is ten feet deep with a three-mile current, in the autumn is only four feet deep with a two-mile current, but in the spring floods goes rushing through its bed forty feet deep with a ten-mile current, it makes a mighty difference to the towns and villages all the way along. The destructiveness of a flood lies in the top few feet of water. In the second place, the navigation of a stream can only be estimated by its lowest depth recorded, and its horse power in the same way. But this same river, which in the autumn was only four feet deep and developed a corresponding horse power, would have an average depth of eight feet with four times the horse power. If then, the water that wastefully and ruinously flows down in the spring is conserved all through the summer, the river has been made more than four times as valuable." "And how is this done?" "That's too big a subject to take up now. Still, you can understand that if you dam the stream high up, and divert all the water over a certain height into immense reservoirs, the water could be let down gradually later. But that all depends on the measurement, which is taken daily for years, often--as in the case I was in--from a cable stretched from bank to bank, from which a little 'bos'un's chair' is hanging on a pulley, so that sitting in this little framework you can reach up to the cable and pull yourself to and fro. The one over the Susquehanna, where I was, is over a mile long, and of course it's pretty high up to allow for the sag, which is not small on a wire of that immense span." Roger had a host of questions to ask but kept silent, not wanting to monopolize the talk when older men were there.