manifestly possible: otherwise it would not have happened. Still there
are even some tragedies in which there are only one or two well known
names, the rest being fictitious. In others, none are well known, as in
Agathon's Antheus, where incidents and names alike are fictitious, and
yet they give none the less pleasure. We must not, therefore, at all
costs keep to the received legends, which are the usual subjects of
Tragedy. Indeed, it would be absurd to attempt it; for even subjects that
are known are known only to a few, and yet give pleasure to all. It
clearly follows that the poet or 'maker' should be the maker of plots
rather than of verses; since he is a poet because he imitates, and what
he imitates are actions. And even if he chances to take an historical
subject, he is none the less a poet; for there is no reason why some
events that have actually happened should not conform to the law of the
probable and possible, and in virtue of that quality in them he is their
poet or maker.

Of all plots and actions the epeisodic are the worst. I call a plot
'epeisodic' in which the episodes or acts succeed one another without
probable or necessary sequence. Bad poets compose such pieces by their
own fault, good poets, to please the players; for, as they write show
pieces for competition, they stretch the plot beyond its capacity, and
are often forced to break the natural continuity.

But again, Tragedy is an imitation not only of a complete action, but of
events inspiring fear or pity. Such an effect is best produced when the
events come on us by sunrise; and the effect is heightened when, at the
same time, they follow as cause and effect. The tragic wonder will thee
be greater than if they happened of themselves or by accident; for even
coincidences are most striking when they have an air of design. We may
instance the statue of Mitys at Argos, which fell upon his murderer while