WHAT OUGHT AN ARTIST TO DO? 65 



the under surface of the hind feet upturned, is really seen 

 by us all every day in the dog, and is recorded in instan- 

 taneous photographs of that animal going at full speed. 

 In fact, the gallop of the dog (and of some other small 

 animals) is a series of jumps ; the animal " bounds along." 

 But this is a totally different thing from the gallop of the 

 horse. It is probable that the dog's gallop was trans- 

 ferred, so to speak, to the horse by artists, and a certain 

 justification for it was found in one of the attitudes of a 

 jumping horse, which, however, never exhibits both the 

 front and the hind legs simultaneously in so completely 

 horizontal a position as they are made to take in the 

 Mycenaean gold-work and the modern " racing plates." 



How, then, we may now ask, ought an artist to repre- 

 sent a galloping horse ? Some critics say that he ought 

 not to represent anything in such rapid action at all. But, 

 putting that opinion aside, it is an interesting question as 

 to what a painter should depict on his canvas in order to 

 convey to others who look at it the state of mind, of 

 impression, feeling, emotion, judgment, which a live, gallop- 

 ing horse produces in him. The scientific draughtsman 

 would, of course, present to us a series of drawings exactly 

 like the instantaneous photographs, his object being to show 

 what " is," and not what the artist aims at, namely, what 

 " appears," " seems," or (without pondering and analysis) 

 " is thought to be." The painter, in his quality of artist, 

 would be wrong to select any one of the dozen or more poses 

 of the galloping horse published by Muybridge, each limited 

 to the fortieth of a second, since no human eye can fix (as 

 the photographic camera can) separate pictures following 

 one another at the rate of twenty a second, each enduring 

 one fortieth of a second, and each separated by an interval 

 of a fortieth of a second from the next. All the phases 

 which occur in any one-tenth of a second (only two, or 

 possibly three of the Muybridge series shown in PI. I) are 



5 



