step," and not "two tracks," because we always begin by going to one side, which is far from the two tracks.

When a horse does the work I have just described, it is said that he goes sideways, no matter how bad may be the position of his head and neck. But for this work to merit being called " work on two tracks," the position should be correct. Yet the position is correct, and a horse moves truly on two tracks when he goes obliquely forward on two parallel lines, the one made by the forehand, the other by the hind quarters. He thus advances from the side, with the head and forehand always leading the hind quarters. The head and neck ought to be kept high and slightly bent to the side towards which the horse is going. Above all things, he should be well in hand, light, and in a cadenced pace.

In my opinion, this is the longest and most difficult work. If we seek to obtain it complete and correct from the beginning, we will either get no good result, or we will cause the horse to at once resist, because he had not passed through the " mill " of lateral, direct and diagonal efforts which I have described.^[1]

My reason for laying considerable stress on the subject of " two tracks," is that this work has a great influence on all the subsequent breaking, where we always meet during propulsion the same combined action of the diagonal " aids."

When the horse knows how to go on two tracks, he will seek every means to escape from the rassembler. At first, he refuses to yield to the direct leg ; he then yields too much