Therefore he must not aim at accuracy of reproduction of natural fact nor even of visual fact, but at the transference to another mind of his own mental condition--his inner judgment as to "things seen"--by means of necessarily imperfect pictorial mimicry. He must therefore avoid startling or abnormal truthfulness of observation of the unessential and even more strictly must he refuse to make his picture a scientific diagram demonstrating what "is" rather than what is "seen" or is "thought to have been seen." On these grounds I find that the most satisfactory pictures of the galloping horse are those which combine a phase of the movement of the front legs with a phase of the movement of the hind legs, not simultaneous in actual occurrence, but following one another. It is for the artist to select the combination best suited to producing the mental result aimed at. Some of the Chinese and Japanese representations of the galloping horse and some of their European imitations (but not all--certainly not that of Stubbs, of the Epsom Derby of Gericault, and the racing plates) seem to me to be eminently satisfactory and successful in this respect. In the pictures to which I allude (Pl. III, figs. 3 and 5) all the legs are off the ground; the front legs are advanced, but one or both may be more or less flexed, whilst the hind legs, though directed backwards with upturned hoofs, are not nearly horizontal (as they actually are in the galloping dog), but show the moderate extension which really occurs in the horse, and is recorded by instantaneous photography. This pose, favoured by many European and Japanese artists, can be obtained by uniting the outstretched hind legs of fig. 9 of the Muybridge series (Pl. I), with the outstretched forelegs of fig. 6, as shown in Pl. I, fig. 12, or by uniting the hind legs of fig. 10 with the forelegs of fig. 4 as shown in Pl. III, fig. 1. With regard to the representation of other "gaits" of the horse than that of the rapid gallop--such as canter, trot, amble, rack, and walk--I have no doubt that instantaneous photography can (and in practice does) furnish the painter with perfectly correct and at the same time useful and satisfactory poses of the horse's limbs. These, though of longer duration than the poses of the gallop, can only be correctly estimated by the eye with great difficulty, and only sketched by artists of exceptional skill and patience. The movement of the wings of birds in flight has been very successfully analysed by instantaneous photography. Some of the poses revealed must familiarise the public with what can be, and, in fact, has been, observed in the case of large sea-birds, by the unassisted eye, and has been represented in pictures by the more careful observers of nature among modern painters. A large sea-bird sailing along with apparently motionless wings has been photographed in the act of giving a single stroke so rapid as to escape observation by the eye.