one after another, the succession of steps being as follows : off hind, near hind, off fore, near fore.

In the full speed gallop this fact is well marked, but the speed of the pace prevents it being apparent. In the stationary canter, and still more in the canter to the rear, it is much less marked, and for this reason it is almost imperceptible. In fact, at this pace the feet only graze the ground, and are drawn back for only a few inches. Instruments of precision or instantaneous photography would be required to prove this in both cases, but the fact remains. We can therefore say that in the full speed gallop and in the canter without gaining ground and to the rear, there are four successive impulsions.

Let us note, however, that in the canter without gaining ground, and in the canter to the rear, the diagonal acts in the same way as in the ordinary canter, with the single difference—which is a point I wish to bring to light—that the support of the hind leg precedes the support of the fore leg of the diagonal by an extremely short though actual interval.^[1]

To obtain the canter without gaining ground, I begin by shortening the canter every day while maintaining the propulsion—that is to say, by pressing the horse with the legs up to his bit, the play of which should be fine in proportion to the extent of the rassembler. By gradually decreasing every day the length of the strides, I obtain the canter without gaining ground with impulse, but not with the horse getting behind his bit. When the animal is behind his bit in the canter, it is impossible to keep him up to his bit, and he necessarily gets away from his rider by reining back.