In the stade-race, the runners were divided into heats (τάξεις), which were drawn by putting into a helmet lots marked with the different letters of the alphabet.[491] If, as seems probable at Olympia, each heat consisted of four, there would be four lots marked A, four B, and so on. It appears that there was no second draw, but that all the winners of heats ran together in the final, so that the final winner had won twice.[492] The starting lines at Olympia provided room for twenty runners at once. In short races the field is often a large one, and we hear of no less than seven Crotoniats winning their heats in a single Olympiad.[493] There is no reason for supposing that the heats were always limited to four. The number would naturally be determined by the number of entries and the length of the starting lines. On the Panathenaic vases representing this race we find usually four, but sometimes two, three, or five runners taking part, though it is unsafe to draw conclusions from this evidence, the number of figures being largely determined in a drawing by considerations of space.[494] Of one thing we may feel sure in spite of assertions to the contrary in modern text-books: the heats were so arranged as to avoid the necessity of a bye. A single odd runner would be attached to one of the heats, two or more would form a heat by themselves. Whether heats were employed in the longer races we have no evidence to determine. The runners were separated from one another at the start by posts placed in the stone sill, and in later times by more massive pillars, and similar posts or pillars marked the finish or turning point. It has been suggested that ropes were fastened to these posts, which ran the length of the course and separated the runners from one another. Unfortunately, there is no evidence for this natural and attractive explanation, and such evidence as we do possess is unfavourable. We hear occasionally of runners interfering with one another by holding, tripping, and running across.[495] Such foul practices seem to have been rare, and were of course strictly forbidden. The competitors at Olympia swore a solemn oath to abstain from all foul play. But on a roped course these practices are impossible. They may, of course, have been confined to the long-distance races, in which the course was certainly not roped. But this is mere supposition, and in the dearth of evidence we must look for some other explanation of the posts. In the first place they must have served as guide-posts for the runners, a very necessary aid in a broad track 200 yards long,—in which it is by no means easy to run straight without assistance. Possibly each post was distinguished by some special sign or colour. Then in the diaulos the runners probably turned round these posts, each turning round his own post. Finally, as the use of the tape seems to have been unknown at the finish, they must have given the judges a most useful line for judging a close finish.