830 aeistotle's treatise on poetry. the length of each performance must be regulated by the hour-glass'. But, if we determine this measure by the nature of the thing itself, the more extensive the fable, consistently with the clear and easy compre- hension of the whole, the more beautiful will it be, with respect to magnitude. — In general, we may say, that an action is sufficiently extended, when it is long enough to admit of a change of fortune from haj)py to unhappy, or the reverse, brought about by a succession, neces- sary or probable, of loell-connected incidents. Cap. viir. ^ pig^ ig j^QJ3 Q^^Q^ g^g some conceive, merely because the hero of it is " one. Tor numberless events happen to one man, many of which are such as cannot be connected into one event; and so likeMdse, there are many actions of one man which cannot be connected into any one action. Hence aj)pears the mistake of all those poets who have com- posed Herculeids, Theseids, and other poems of that kind. They con- clude, that because Hercides was one, so also must be the fable of which he is the subject. But Homer, among his many other excellencies, seems also to have been perfectly aware of this mistake, either from art or genius ; for wlien he composed his Odyssey, he did not introduce all the events of his hero's life, such, for instance, as the wound he received upon Parnassus; his feigned madness when the Grecian army was assembling, &c. ; events not connected, either by necessary or probable consequence, with each other; but he comprehended those only which have relation to 07ie action, for such we call that of the Odyssey. And in the same manner he composed his Iliad. As, therefore, in other mimetic arts, one imitation is an imitation of one thing, so here the fable, being an imitation of an action, should be an imitation of an action that is one and entire; the parts of it being so connected, that if any one of tliem be either transposed or taken away, the wJtole will be destroyed or changed; for whatever may be either retained or omitted, without making any sensible difference, is not j^roperly a 2^<^Tt. Cap. IX. It appears further, from what has been said, that it is not the poet's Trawdy to province to relate such things as have actually happened, but such as relight have happened; such as are possible according either to probable or necessary consequence. For it is not by writing in verse or prose that the historian and the poet are distinguished : the work of Herodotus might be versified, but it would still be a species of history, no less ^ We have here in the original the unmeaning addition, Cicnrep Trork Kai &XKoTe <paalp.—J. W. D.
