performance, when an actor comes in, the question as to whom he shall see first, and which person he shall address first, is decided, not by the convenience of the poet, but by the science of optics. The experience of the modern stage is sufficient to prove that this is not the case. It would be unsafe then to lay any stress on the instances just cited. The cases in which there are adequate grounds for supposing that the actors entered or departed by the orchestra amount to no more than about thirty. The question is whether these cases are sufficient to justify a wider inference. Are we to assume that, because the actors sometimes used the parodoi, they did so always? On the one hand it may be said that in the early theatre, with its low stage and easy communication between stage and orchestra, there was nothing to stand in the way of such a practice. On the other hand there is the fact that in the later Greek theatre the actors, when coming from a distance, usually entered by the side-wings.[604] Of course in this later theatre, with its twelve-foot stage, there were obvious reasons for doing so. Still, the existence of the practice in late times is a presumption in favour of its existence previously. Moreover, when side-wings had once been introduced, nothing could be more natural than to use them as entrances. The convenience to the actors would be very great. It is difficult to see why they should have been compelled to go all round by the parodoi when there was an easier mode of entrance close at hand. On the whole, therefore, it seems most probable that the side-entrances were generally used by the actors even as early as the fifth century, and that the orchestra was only employed in special cases, such as processions with the chorus.