Such is the literary evidence. No single item of it is decisive for an early occupation. Those which are definite can be questioned on point of fact. Sappho, whose evidence alone cannot be so questioned, may conceivably have written after 570. Combined however they make it probable that Naukratis rose to importance before the days of Amasis, and that Herodotus either confused the foundation of the city with that of the Hellenium[546] or else did not intend his words “gave the city for Greeks who came to Egypt to dwell in” to imply that the Milesians were not there in force before the granting of this concession[547]. Nevertheless, if we were limited to these literary sources, we could not be certain that where Diodorus seems to supplement his predecessors he was not merely adding details that they appear to imply. That is in fact the view of his narrative that some modern scholars apparently hold[548]. Even if this were so, his additions would still have a certain value. If Diodorus, writing in the first century B.C., read between the lines of Herodotus the same unstated implications that have been read there in recent times, the coincidence points to the probability that this reading is not altogether wrong[549]. [Sidenote: (_b_) archaeological evidence:] That is as far as the texts take us. For further light we must look to archaeology. The new light began by increasing the perplexity. Petrie and Gardner both claimed that their excavations at Naukratis proved that it had been an important Greek city from the middle of the seventh century. But their main arguments were before long shown to be mistaken, and somewhat naturally it began to be assumed that they must be wrong in their conclusions. [Sidenote: (i) excavations of Petrie, who dated Naukratis from the time of Psammetichus I;] Petrie[550] based his arguments on the following observations. In the South part of the town he came across what he described as a scarab factory. There were numerous scarabs of Psamtek I, some of Psamtek II, and some that are probably of Apries; but none of Amasis. This seems to date the factory from well before 610 till after 589. Two feet beneath the factory was a burnt stratum of plain potsherds which must take us back a good way further, to at least 650 B.C. and probably earlier. The scarabs are imitation Egyptian and are taken by Petrie to be Greek. Further South, but also within the area of the burnt stratum, there is a large enclosure which he describes as surrounded by a strong brick wall. This he identified with the Hellenium, where Herodotus states that nine Greek cities had quarters assigned to them by Amasis. The dimensions of the bricks point to the early Saite period. [Sidenote: (ii) further excavations by Hogarth invalidated Petrie’s arguments;]