Professor E. A. Gardner also shows reasons of a technical character in favour of the earlier date.[437] The foundations of the chryselephantine statue of Dionysus by Alcamenes are of conglomerate and breccia. Alcamenes was at work during the latter half of the fifth century; and the later temple in the precinct below the theatre was built to contain this statue. Now, as Professor Gardner points out, it is unlikely that the Athenians would have undertaken so costly a work in the later part of the Peloponnesian war. On the other hand, there is no trace of the use of breccia in foundations in the Periclean age. The temple probably therefore dates from the time between the Peace of Nicias in 421 B.C. and the Sicilian expedition in 415 B.C. And if the temple was built then, it is not unlikely that the theatre may have been begun at the same time. The fact that the architectural technique of the theatre, particularly in the use of conglomerate blocks, is the same as that of the temple points the same way. The work may have begun about B.C. 420, and progressed gradually and continuously up to the time of Lycurgus. The exact year in which the higher stage was erected cannot, of course, be fixed.[438] Puchstein also doubts whether the whole of the existing auditorium was built in the time of Lycurgus.[439] There is a fragment of a wall (not marked) in front of _a—a_ in the plan, which Dörpfeld does not mention in his text, though he marks it in one of his plans.[440] This, Puchstein suggests, is the supporting wall of an auditorium older than the Lycurgean. Besides this he finds evidence of stone seats in the fifth century. It has already been mentioned that a stone built into the western wing of the auditorium contains a fifth-century inscription.[441] This inscription consists of the words βολῆς ὑπηρετῶν, and was probably part of a seat-step, reserved for the servants of the βουλή. If so, there must have been a stone auditorium before the time of Lycurgus.