TERMESSUS itself was evidently at the entrance of the defiles whereby Pisidia communicates with Pamphylia and Lycia. Arrian says that “the men of Termessus occupy a site very lofty and precipitous on every side, the road passing close to the city being very difficult, as the mountain reaches down from the city to the road. There is over against this, another mountain not less precipitous, and these form a gate, as it were, on the road,” &c. This statement is fully confirmed by the observation of General Köhler (ap. Leake, Asia Minor, pp. 133-135): “The two great ranges on the west and north of the plains of Adalia,” says he, “now approach each other, and, at length, are only divided by the passes through which the river finds its way. The road, however, leaves this gorge to the right, and ascends the mountain by a paved and winding causeway, a work of great labour and ingenuity.”[77] Alexander the Great, it would seem, despaired of taking the town; or, possibly, thought its siege would detain him too long; he, however, forced the defiles, passing on to the north to Cormasa, Cremna, and Sagalassus, a course probably pursued by Manlius subsequently.[78] CREMNA, where, owing to its great natural strength, the Romans placed a colony (Strab. xii. 569), has been carefully examined by Mr. Davis (“Anatolica,” p. 182), who gives also a plan, showing the construction of this remarkable fortress. His description is as follows:[79] “It (Kremna) is a plateau of limestone, which is bounded on three sides by precipices, some extremely deep and abrupt.” Footnote 77: There is some confusion between the two Termessi, one of which is apparently to the left of the road passing W. and N.W. from Adalia. This we think was _Termessus Minor_—the _Almalu_ of Mr. Davis. The more important place, _Termessus Major_ (on its coins μείζων), was at the head of the pass described. These views are confirmed by Eustath. and Dion. Perieg. v. 858, Stephan. Byzant., and Hierocles. At a later period, the see of Termessus had united with it the churches of two other places—Jovia and Eudocia. Footnote 78: Cramer and some other geographers place Cremna to _the N._ as well as the E. of Sagalassus, where it _could not have been_. Footnote 79: The description in Arundell, vol. ii. pp. 59, &c., shows that he had explored the same ruins forty years before Mr. Davis, under the idea they were those of Selge, though, on his plate, he adds the words, “Acropolis of Germe—Cremna.” Colonel Leake, too, suggested that “Germe” was perhaps a corruption of “Cremna.” Had Mr. Arundell reflected on an inscription he himself copied there ... ΛΔΗ ... ΝΑΤΩΝ, he might have seen that the last word could naturally be supplied as ΚΡΗΜΝΑΤΩΝ—“of the people of Kremna.” Zosimus says the winding path up to the fortress was called by the natives the _Snail_.