There is more difficulty in describing the arrangement of the remainder. Strabo, who has followed the Catalogue in Thessaly with great minuteness, seems to have noticed the circular arrangement: at least he speaks of the κύκλος τῆς Θετταλίας, and the περιόδεια τῆς χώρας[485]. But when he comes to the sixth division, that of Eurypylus, he appears to find it impossible to fix with any confidence the site of Ormenium: and says, καὶ ἄλλα δ’ ἐστὶν ἃ λέγοι τις ἂν, ἀλλ’ οὖν ὀκνῶ διατρίβειν ἐπὶ πλέον[486]. And further on he observes, that the displacements and changes of cities, and mixtures of races, have confounded the names and tribes[487], so as to make them in part unintelligible to men of his day: where we are anew reminded of the passage of Thucydides, in which he tells us, that the most fertile tracts underwent the most frequent changes of population[488]. The δυναστεία of Eurypylus is in our maps commonly placed on the sea coast, but as it appears, with little authority of any kind: while, after all the proof we have seen of continuous arrangement, it seems incredible that, in this instance alone, Homer could have followed an order such that the δυναστεία should not march either with that which precedes, or that which follows, but should be severed from them by a line of territories intervening, which he has already disposed of. To judge from analogy with the otherwise uniform rule of the Catalogue, the dominions of Eurypylus must have been somewhere conterminous both with those of the Asclepiads, and with those of Polypœtes. Waiving however any effort to fix positively their site, we find the other four remaining contingents connected by a zigzag line[489], like that which was used in southern Greece. The leaders are as follows: I. Podaleirius and Machaon (729-33). (Eurypylus 734-7, omitted.) II. Polypœtes (738-47). III. Gouneus (Enienes, Perrhæbi, and Dodona, 748-55). IV. Prothous (the Magnesians, 756-9). In this view Homer appears to subdivide Thessaly into two figures, as he had done Southern Greece: and in both cases one of them is curvilinear, in which the eastern parts are arranged: the other a zigzag, which includes the western portions. I have described this geometrical arrangement, as of great interest in connection with the question, whether the poems were written or recited; and also as it seems to be in itself highly ingenious. It seems to distribute in rude but real symmetry before the eye of the mind, an assemblage of objects between which it would at first sight appear almost impossible to frame any link of connection.