Fig. 5 represents a _circular current_ over the group of three rocks. This is a very deceptive case, for it looks so easy that at first it is likely to be treated carelessly. If the boat were supposed to be a substance floating, but without weight, it would have its direction of motion instantly altered by that of the current. But the boat has weight, and as it has velocity (that of the current even if the boat is not urged also by the paddle so as to have "way" through the water), therefore it will have _momentum_, and the tendency will be to continue the motion in a straight line, instead of a curve guided solely by the current. In all these cases, therefore, it will be found (sometimes inexplicably unless with these considerations) that the boat _insists_ upon passing between _A_ and _C_, where it must not be allowed to go on the hypothesis we have started with; and if it effects a compromise by running upon _C_, this is by no means satisfactory. This class of cases includes all those in which the river makes a quick turn round a rock or a tongue _B_, where the boundary formed by the rock _A_ on the outer bend of the stream is a solid bank, or a fringe of growing trees, or of faggots artificially built as a protection against the erosion of the water. This case occurs, therefore, very frequently in some fast rivers, say, at least, a hundred times in a day's work, and perhaps no test of a man's experience and capacity as a canoist is more decisive than his manner of steering round a fast, sharp bend. The tendency of the canoist in such cases is always to bring the boat round by paddling forward with the outer hand, thereby adding to the "way," and making the force of the current in its circular turn less powerful relatively. Whereas, the proper plan is to back with the inner hand, and so to stop all way in the direction of the boat's length, and to give the current its full force on the boat. Repeated lessons are needed before this is learned thoroughly. The case we have last remarked upon is made easier if either _A_ or _C_ is a breaker, but it is very much increased in difficulty if the rock _B_ is a breaker or is a strong tongue of bank, and so deflects the current outwards at this critical point. The difficulty is often increased by the fact that the water inside of the curve of the stream may be shoal, and so the paddle on that side strikes the bottom or grinds along it in backing.