The first half of the third century was a period of very good work in portraiture, which is, however, a subject by itself. The Demosthenes of Polyeuctes is dated about 280 B.C., and the statues of Aeschines, Aristotle, and others show the existence of an admirable school of portrait sculptors at this time in Athens. But ideal sculpture shows a sad falling-off. The Themis of Chairestratos (Fig. 40) belongs approximately to this period, and it is marked by a great formality, not only in pose but in the treatment of hair and drapery. The classical period of sculpture in Athens was followed by what we must call an academic period. The foreign schools were developing on lines of naturalism, but at home sculptors tended merely to formalize the work of the fourth-century masters, and to produce statues of mechanical correctness without any vitality at all. We have seen the beginning of this tendency in the drapery system of the followers of Praxiteles. It now affects the whole of Attic sculpture. Old types are adopted again and again, until they become purely mechanical. Drapery styles are similarly used up, and the increasing formality of every department stifles entirely the possibilities of originality. The Hermes of Andros (Fig. 39) is a good example of this kind of crystallization of types. The statue was found in connexion with a tomb, and it is clearly a memorial statue. Its companion was a female figure reproducing exactly the pose and drapery of the draped female figure from Herculaneum at Dresden. The date would seem to be late third century. The Hermes itself is a replica of a type known in the Antinous of the Belvedere and other statues, and is a product of the Praxitelean school, like the Dresden figure. But the influence of Praxiteles is not alone in it. We have a clear use of Lysippic proportions and some Lysippic influence in the head. This eclecticism is an invariable mark of archaistic art. The sculptor, who has no new message of his own to deliver, looks back to antiquity for his types, but does not imitate one statue directly. The only form of originality which he is able to use is originality of combination and selection. Consequently he absorbs details from several artists and produces work which we label Lysippo-Scopaic, or Lysippo-Praxitelean, &c. We have seen how the late fourth-century artists in Asia Minor combined characteristics of Scopas and Praxiteles. The late fourth-century and third-century Attic artists made use of all their predecessors, and produced statues in which we can detect the _disiecta membra_ of half a dozen styles. At the same time we may recognize the general predominance of Praxitelean tradition over that of the other artists and a universal predilection for marble instead of bronze.