Again, let us suppose that while the air is absent the force of gravity comes into play, what effect will that have? It will gradually pull the shell downwards out of its horizontal course, making it describe a beautiful curve. But, someone may think, does not a rapidly-moving body remain to some extent unaffected by gravity? Not at all: it falls just the same and just as quickly as if it were falling straight down. If our imaginary horizontal gun were set at a height of sixteen feet and a shell were just pushed out of it so that it fell straight down the shell would touch the ground in one second. If the ground were perfectly flat and the shell were fired so that it reached a point half a mile away _in one second_ it would strike the ground exactly half a mile away. You see, the horizontal motion due to the explosion in the gun and the downward motion due to gravity go on simultaneously and the two combined produce the curve. To make this quite clear, let us imagine two guns precisely alike side by side and both pointed perfectly horizontally. From one the shell is just pushed out: from the other it is fired at the highest velocity attainable: both those shells will fall sixteen feet or a shade more in one second, and if the ground were perfectly level both would strike the ground at the same moment although a great distance apart. Clearly, then, the faster the shell is travelling the more nearly horizontally will it move, for it will have less time in which to fall, and the slower the more curved will be its path, from which we see that the air by reducing the velocity causes the curve to become steeper and steeper as the shell proceeds. If, then, our gun is placed low down, as it must be on a ship, to get the longest range we must point it more or less upwards because otherwise the shell will fall into the water before it has reached its target. When we do that we complicate matters somewhat, for gravity tends to reduce the velocity while the shell is rising and to add to it again while it is falling. We need not go too deeply into that, however, so long as we realize that, whatever the conditions may be, the shell in actual use has to follow a curved course, first rising and then falling. The really important part about a shell's journey is the end. So long as it hits it really does not matter what it does on the way, and if it misses it is equally immaterial. The reason why we need to bother about the first part of the trip is because upon it depends the final result. Whatever the trajectory may be we see that the shell must necessarily arrive in a slanting direction. And the more steeply slanting that direction is _the less likely is the target to be hit_.