To many persons the undisputed fact that the Homeric poems were composed to be recited, not read, has appeared a convincing proof that they could not have originally assumed the form in which they are known to us. For setting aside the difficulty of preserving by the aid only of memory, and the still greater difficulty of _composing_ a long poem without help of the manuscript, to keep _secure_ the part already completed, what motive, it has been said, could induce the poet to undertake so great and so superfluous a labour? Why indite a poem so much longer than could be recited on any one occasion, and which, _as a whole_, could never be appreciated? But we would suggest that it is not necessary to suppose that the poet commenced his labours with the project in view of writing a long epic, in order to believe that we possess these two great poems very nearly in the original form in which they were composed. If it were the task of the poet or poets to supply a number of songs on the adventures of a popular hero, or the achievements of some famous war, such number of songs _must_ assume a certain consecutive order, the one will necessarily grow out of the other. Let any one reflect for a moment how the work of composition proceeds, and he will perceive that it would be impossible for a poet to take any one such subject as the siege of Troy, or the return of Ulysses, as the theme for a number of separate poems, and not find that he was writing, with more or less continuity, one long entire poem. This continuity would be improved and especially attended to, when a certain _order_ came to be preserved (as we know it was) in the recitation of the several poems. We have no difficulty, therefore, in believing that, in the time of Pisistratus, the _editors_ of Homer might have had very little to do to give them that degree of completeness and unity which they at present display. A number of consecutive songs upon the same subject would naturally grow into an epic.