The same commentators believe that Pausanias’ story of Milo bursting a cord drawn round his brow by swelling his veins arose from the victor band on the statue, and the story of the strength of his fingers from the position of the fingers on it. [Illustration: FIG. 15.—Bronze Statuette of a Palæstra Victor, from the Akropolis. Akropolis Museum, Athens.] We have seen in the “Apollo” statues a considerable variety of physical types. In the sixth century B. C. the artist was feeling his way and was hampered by local school tendencies. At first he knew only how to produce rigid statues in the conventional Egyptian attitude with the arms glued to the sides, the two halves of the body being symmetrical and the hips on the same level. He gradually improved on this model, making the position more elastic—as in the statue of Milo—rightly indicating bones and muscles and giving to the figure natural proportions. Bulle has shown on one plate[859] three statuettes which illustrate the improvements reached in bronze in various parts of Greece by the end of the sixth century B. C. To the left is represented a victorious palæstra gymnast—as is indicated by the remnants of akontia in the hands—in the Akropolis Museum (Fig. 15);[860] in the center is the Payne Knight statuette of the British Museum,[861] carrying a fawn in the right hand, which is a copy of the _Philesian Apollo_ which stood in the Didymaion near Miletos; to the right is Hermes with the petasos, short-girded tunic, and winged sandals, holding a ram in the left and probably a kerykeion in the right hand.[862] The attributes of the three, then, attest respectively a victor, Apollo, and Hermes. In all three the arms are freed from the body, and the muscles of the breast, chest, and abdomen are indicated, though carelessly in the case of the victor. The proportions of the three vary greatly; the Attic victor has a large head, broad shoulders, powerful chest, long body, and short legs; the _Apollo_ has long legs, shorter though slimmer body, and small head;[863] the _Hermes_ has a clearly outlined figure and shows the careful modeling so characteristic of the schools of Argos and Sikyon in the fifth century B. C. Bulle shows that the further development of the “Apollo” type was halted by the Argive school, which, while continuing the restful pose of these figures, counteracted their rigidity by inclining the head to the side and throwing the weight unevenly on the legs by lowering one hip and further advancing one foot. The central line was no longer vertical, but curved, and it was now possible to give greater detail to chest and abdomen. Polykleitos finally perfected this curve and threw back the left foot, resting the weight of the body on the right—from which time on we have the regular scheme of “free” and “rest” legs. Despite all these later improvements, Olympic victors continued to set up statues in the rest attitude of the “Apollo” type down perhaps into the third century B. C.