_Her_. How? Why, two of them must have alpha, two beta, and of the next two pairs one has certainly drawn gammas and the other deltas, so that four letters have been used up over eight competitors. Obviously, then, the next letter, which is epsilon, is the only one that can be odd, and the drawer of it is the bye. _Ly_. Shall I extol your intelligence, or would you rather I explained to you my own poor idea, which differs? _Her_. The latter, of course, though I cannot conceive how you can reasonably differ. _Ly_. You have gone on the assumption that the letters are taken in alphabetical order, until at a particular one the number of competitors runs short; and I grant you it may be done so at Olympia. But suppose we were to pick out five letters at random, say chi, sigma, zeta, kappa, theta, and duplicate the other four on the lots for eight competitors, but put a single zeta on the ninth, which we meant to indicate the bye--what then would you do if you came on the zeta first? How can you tell that its holder is the bye till you have been all round and found no counterpart to it? for you could not tell by the alphabetical order, as at Olympia. _Her_. A difficult question. _Ly_. Look at the same thing another way. Suppose we put no letters at all on the lots, but, instead of them, signs and marks such as the Egyptians use for letters, men with dogs' or lions' heads. Or no, those are rather too strange; let us avoid hybrids, and put down simple forms, as well as our draughtsmanship will allow--men on two lots, horses on two, a pair of cocks, a pair of dogs, and let a lion be the mark of the ninth. Now, if you hit upon the lion at the first try, how can you tell that this is the bye-maker, until you have gone all round and seen whether any one else has a lion to match? _Her_. Your question is too much for me.