I must begin with an explanation of my reasons for writing this paper. I have for some years been interested in modern attempts to reconstruct ancient campaigns and battles—especially those between Greece and Persia—in fact, most of the arguments I am using tonight were first written down in a rather different form in 1913. But I should not have thought it worth while to read them to this Society had not Admiral Custance's book, his lectures, his address to this Society and the discussion which followed it made me feel that this subject is still one of fairly general interest, that we are still far from arriving at certainty with regard to the history of ancient fights, and that it may be worth while to raise some general questions such as, ‘How far it really is possible to reconstruct ancient battles with any finality’ and ‘how far the methods of attempting to do so usually followed by modern writers really are the soundest methods to employ’.