do their high school work without taking off their bridles. Here, also, I found their respective mouths to be exactly the same as before.

While hacking, these horses were light in hand, without being on their haunches.^[1] On the race-course they pulled just as much as if they had never been taught high school work ; and, having returned to the school with their hind legs well under them, they were as light in hand as if they had never been galloped on a race-course.

We should note that in the experiments which I have just described, the respective weights, with these horses, were distributed in three absolutely different ways.

1. When going to the course it was distributed in the proper manner for hacking, and was the result of an ordinary feeling on the reins.

2. During the gallop, the preponderance of weight was on the forehand.

3. In the manège, on the resumption of high school work, the preponderance was on the hind quarters. Consequently, there were three modifications in the distribution of weight. The application of the " aids " were also entirely different, according to the nature of the work. But the essential point on which I differ from Baucher is, that a horse's mouth remains the same, no matter what is the work, use of the "aids," or distribution of weight. In the three distributions of weight I found the same difference of mouth.

I repeated this experiment with twenty thorough-breds which had been broken for high school work, and they all brought me to the same conclusion : that the nature of the mouth does not vary according to the distribution of weight.

It is sometimes evident that a horse which holds his head low makes the rider carry his head and neck ; but if the rider