Carried upon the bridge is a platform which bears a number of instruments, chief among which is a self-recording dynamometer. Now a dynamometer is an instrument for measuring the force of a "pull," and when we call it self-recording we mean that it automatically takes a record of a series of pulls or of a varying pull. In this case there projects below the bridge a lever, to the end of which the model under test is attached. As the bridge rushes along it pulls the model through the water by means of this lever, and the force which is expended in doing so is recorded in the form of a wavy line upon a sheet of ruled paper. If the model slips through the water very easily there is little pull upon the lever and the line drawn by the pen of the instrument remains low down upon the chart. If, however, much power is needed and the pull is a strong one the pen moves and the line rises towards the top of the paper. Any change, whether increase or decrease, is thus shown by the rise or fall of the ink line. One model can be thus tried at various speeds and its behaviour noted under different conditions. Other matters can be investigated too, such as whether or not the bow rises in the water or falls when the boat is in motion, also how much such rise or fall may amount to. The suitability of a certain shape of vessel, moreover, can to a certain extent be seen by observing the commotion which it makes in the water. Everyone has noticed the way in which a ship throws up a wave at its bows, and that bow-wave, as it is termed, represents so much energy being wasted. The power of the engines is absorbed to a certain extent in making that wave. It is impossible to make anything which when forced through the water will not make some wave, but certain forms cause less of it than others, and the designer of a ship seeks to find that form which will make the smallest bow-wave. In like manner the eddies which a ship leaves in its wake are the result of wasted energy, and the ship must be so shaped that they too will be reduced to a minimum. Shipbuilders find that there are three things which retard a ship's movement: skin friction, or friction between the water and the sides of the ship; wave making at the bow and eddy making at the stern. The first depends largely upon the smoothness of the ship's surface, the second and third depend upon its shape. If a model behaves badly in the tank the fault may be either too much wave making or too much eddy making, and which of these it is the dynamometer does not of course tell. In many cases the experienced eye of the tank officials furnishes the clue to the trouble, but in some cases a cinematograph is used to make a complete series of photographs of the model and the water around it as it rushes from end to end. These can then be studied in conjunction with the chart and the cause of the fault discovered.