The theory of Puchstein, already so often alluded to, ascribes to Lycurgus the construction of the proscenium consisting of stone columns and pinakes, and throws back to the end of the fifth century the Lycurgean structures usually so called. His principal ground for this change of date lies in the development which he traces in the form of the columns in question.[435] He thinks it certain that the use of full columns must have preceded that of half-columns, and that columns without special contrivances for holding pinakes must be earlier than simple ones. Thus the full columns of the proscenia of Athens, Sicyon, and the Peiraeeus, which have no such contrivances, will belong to the earliest period of stone proscenia; they will be earlier than those of Megalopolis and Eretria, which have rims for holding the pinakes, and still earlier than the plain half-columns of Epidaurus and the grooved half-columns of Priene, Assos, Delos, Pleuron, Oropus, &c. The proscenia of Priene, Pleuron, and Delos appear to belong to the third century B.C.; and Puchstein accordingly throws back the Athenian columned proscenium to the latter half of the fourth century, the time of Lycurgus. The theory is at least plausible; but it is not certain. Development is not always in a straight line or in logical order, and does not always require intervals of many years between one stage and another; different experiments may be tried simultaneously in different cases, and recurrence to old types, or preservation of them after new ones have been invented, is a common thing in the history of architecture. The form of the proscenium, therefore, cannot be used with certainty as a chronological criterion, though it may be very suggestive. It follows, in Puchstein’s view, from the earlier dating of the stone proscenium, that the so-called Lycurgean stage-building, with its deep side-wings, must have been erected some time before Lycurgus, towards the end of the fifth or beginning of the fourth century. The date which Puchstein suggests for the stone proscenium at Athens is certainly more probable on _a priori_ grounds than that given by Dörpfeld. According to Dörpfeld’s chronology, the earliest stage-buildings at Athens were apparently later than those at Eretria and other Greek cities. Dörpfeld has conceded that the old skene at Eretria is of the fourth or fifth century, and may be older than the Lycurgean.[436] But it is hard to believe that the city in which the drama was first developed should not have been the first also to provide itself with a permanent stage.