¹ The correctness of Mr. Hinke's opinion is evident. He points out that Augustus is represented mourning on a paste in Vienna (imitation turquoise): Delbrueck, , Antike Porträts, Pl. 59, 4Google Scholar.

² This was stated only in conversation at the time, and never printed.

³ Amyntas was killed in the summer of 25: no campaigning in the high Taurus is possible except in the dry season. News would not reach Rome for nearly two months; the defeat of a barbarian king would not be announced pinnatis nuntiis. However quickly plans were made, time was needed for the arrangements, the selection of veterans and their transport to Ephesus, and 350 M.P. inland from the harbour of Ephesus.

⁴ also occurs. remains unexplained; possibly it is for “look,” calling attention to the advertisement on reverse. The spelling is late everywhere.

⁵ In the Transactions of the Amer. Philol. Assoc., LVII, 1926, p. 229, Professor D. M. Robinson publishing this tablet suggests as alternative interpretation Ποτα Μηνη, a female form of the god Men: such an interpretation would be false to the religious idea. Had it been possible, it would be a welcome example of a form Μήνη, Greek equivalent of Manna, the native Anatolian feminine of Mannes.

⁶ οὐδέ is used for οὔτε; Papias did not know Greek well.

⁷ The vocalisation Sīblia, Sūblaion is noteworthy. Enthusiasm for Augustus and Gaius Caesar prompted Kallikles, a wealthy citizen, to pay for coins. Again, in 212, loyalty induced Menodotos and Ailiano (his wife) to pay for a considerable issue of coins with types Zeus, Athena, Herakles, Hermes, Dionysos, Tyche.

⁸ The analogy was indicated to me by Professor Calder many years ago. Worship of the Beast and of his image are spoken of as different: Προσκυνϵῑν τῆ ϵἰκόνι, is a distinct cult (Revelations, XIII, 15Google Scholar; XIV, 9, 11; XX, 4); the Image has breath and speech (XIII, 15), as if a personal being.

⁹ This punctuation, rejected by modern editors, seems infinitely superior to the text preferred by them.

¹⁰ is masc. and fem. (L. and S.), Hesychius has with synonym In the passages of Matthew and Mark, is wrongly rendered dogs. Dogs under the table would be an outrage on Eastern feeling: dogs were unclean animals; but puppies were permitted. Probably Christ said throw to the dogs outside the house; and the woman replied this makes an incident true to life in any village or humble home in Galilee or Asia Minor.

¹¹ The double name was usually stated with father's name, after the first and before the second name, and a signum was added with e.g. Buckler, Inscr. Sard. No. 142.

¹² It cannot be regarded as a true praenomen assumed with the nomen as part of the Roman name: the formula ὁ καὶ is decisive against this.

¹³ Kubitschek, , de Rom. Tribuum origine (Abhandl. III des A.-E. Seminars, Wien, 1882), p. 115Google Scholar. The inhabitants of towns to which the Emperor gave civitas were assigned to his tribe, Julii in Fab., Claudii and Flavii in Quir., Ulpii in Papir., Aelii in Serg.

¹⁴ Professor Calder prefers ‘settlement’ to ‘town’: κώμη is the Greek word used by Strabo.

¹⁵ He corrects my statement that the town ‘is now uninhabited.’ I might have said ‘when I knew it, 1890–1909.’ There was an old mosque in the ‘settlement,’ the only building in 1890–1909.

¹⁶ Obviously the epitaph alludes to the town at which the grave is, not to some distant place. Calder admits that this ‘is probable.’ It may be taken as self-evident.

¹⁷ In Cyprus at second hand since 1878.

¹⁸ Similar is the history of Pisidai, Pisidike, Pisidia, (Klio, 1929, p. 376Google Scholar).

¹⁹ He is our main authority for the Homanadensian War in 10–7 B.C., and a useful subsidiary for Servilius's campaigns in 77–75.

²⁰ Strabo's statement about the range of vision from Olympos fort in Lycia (captured by Servilius) would be correct if the word πᾶσα were omitted; the four countries are visible but not the whole of any. The exaggeration was easily made and is easily removed.

²¹ Cichorius, , P.-W., s.v. 1251Google Scholar, infers that an Ala I Mil. was probably stationed in Palestine both about 100 and 400; and he is supported by the Antiochian inscription.

²² Over the name of the Platea there was some ornament or symbol of bronze: the hole in the stone where it was held is empty.

²³ Professor Robinson always calls the monument Propylaea, for what reason I know not.

²⁴ Perhaps [et] liberis eius should be taken as dative.

²⁵ My correspondents, after using Yalvac in the modern spelling, have now settled on Yalavac as correct.

²⁶ Both Attalos I and II were sons of Attalos.

²⁷ Anatolia may be used as a synonym for Asia Minor: this is not quite correct; it excludes the country south of Taurus, i.e. Karaman, Karamania.

²⁸ Is Κύβελα the Semitic Gebel, Djebel? Sonny, in Philologus, 1889Google Scholar, suggested that Κυβέλη = Ῥέιη = Ὀρείη but this found no support.

²⁹ An example of religious continuity is the cave of St. George of the Limnai, in the form of a roughly hewn chapel in the rock, beside a great natural gateway over the northern Limna in Pisidia: close to the cave is a small old Phrygian monument of simple type. Τρῶγλα was Anatolian, in Τρουγλεττα, Τρωγλοδύται, etc. So also θάλαμος.

³⁰ We made a rapid journey from Serai-Keui railway station to the Phrygian Kings' monuments about Ballenaion and Metropolis in 1887, passing Dionysopolis on our way.

³¹ My copy indicates (4) 1. 3 as separate from (4) 1. 1, which is not on an architectural fragment, whereas (10) and (4) 1. 3 are on such fragments. The architectural measures are slightly different, owing to poor workmanship, but the simple style is the same, and only measurement shows difference. The two fragments were on different sides of the Temple. My copy puts (4) 1. 3 immediately after (4) 1. 1, but marks distinction between them.

³² This arrangement conjoins the names of the dedicator and the Goddess-Empress for whose salvation the dedication is made, and was evidently made purposely. As a worshipper approached he saw the names over the door: the rest of the inscription was engraved in a single line round the other sides of the small Temple. The dedicator may be assumed to be the priest and νεωκόρος of the goddess: see Buresch, , Aus Lydien, p. 18Google Scholar.

³³ The hill is the projecting edge of the Plateau.

³⁴ For example (mason-carpenters), (makers of delicate ware?), and one or two others.

³⁵ A new foundation practically always means a new site adapted to change of social and political conditions. This I have called Hirschfeld's, Canon in H.G.A.M., p. 91Google Scholar, because he first noticed it at Seleukeia-Agrai, though he did not generalise it.

³⁶ Since I wrote on this region, in many cases immediately after I published, many writers have treated it, and to their books I am greatly indebted—Cichorius and Judeich, Head and Imhoof, Radet, Rostovtseff, Buresch, Anderson, etc. The agreement of Imhoof and Head is to me final in numismatic points, and very weighty in the relation of numismatics to history. The ready help of the officials in the British Museum has been invaluable to me since 1879, and is still the same.

³⁷ The punishment was disease, very often fever (an unseen divine fire). Confession and expiation must be made, when she gave warning thus: the punishment was the warning.

³⁸ is a Phrygian word meaning hence is derived [and we may add, High-Landers, and the High Place]. The lower zone, earth, should always be doing what is being done in the high zone, heaven.

³⁹ In V, 113 he assigns the low Maeander valley to Caria (pervagatur postremo Cariam placidus, etc.), but in V, 110 to Lydia (perfusa flexuosis Maeandri amnis recursibus): these two passages seem to be taken from one authority and to imply that the river flows between Lydia and Caria.

⁴⁰ The description is evidently earlier than Agrippa's survey (12 B.C.). Pliny's authority was some Greek geographer; and if we could suppose that Eumenetica is a later insertion for Peltena, Dikaiarchos is probably the authority: he is mentioned in Pliny's list of authorities both in V and in VI.

⁴¹ It is better to count Soublaion as part of regio Eumenetica.

⁴² Hence Κόνιον, Conium, ranked as one of the celeberrima oppida of Phrygia with Kelainai, Andria, Colossae, Carina, Ceraine: V, 145: See Byzantion, V, pt. 3.

⁴³ On the spelling see note 45.

⁴⁴ Twelve cities on the western seismic zone suffered in A.D. 23. Kibyra (etc.) suffered a year later; all were included in the imperial relief and on the monuments at Rome (lost) and Puteoli. Rhodes suffered under Pius, and Laodicea, etc. a year or two later in 151–2.

⁴⁵ Compare Kastaballa-Hieropolis of Cilicia, where double Λ occurs rarely; but evidently the second element was bāla, balla.

⁴⁶ Compare Plautus's spelling Ludus, Surus, Tuestes.

⁴⁷ In Röm. Mitt., 1900, p. 200 f., Meurer pointed out that there is no nipple in the ‘mammae,’ and that the want is extraordinary, as the goddess is the nourishing power: he explained the want as caused by a robe; but the figure is nude. Ova, not mammae, are represented.

⁴⁸ The relief is from Sinda (Istanoz) in Pisidia (Konia Vilayet included Sparta and district. It was cut in 1922). The relief here is, as Buckler sees, the Mother Goddess: but Mendel calls her Helena. The horsemen are evidently Kabeiroi.

⁴⁹ In Judeich No. 195, money to be given to persons whose names appear in the schedule or list: this list would be attached to the antigraphon deposited in the Archives.

⁵⁰ should be should be

⁵¹ Read Αναις for ἀν. [δ]ίς.

⁵² The published text has in 4, and in 8, (explained as an unknown word). The word should be restored in an inscription of the Axylon, M.A.M.A., I, No. 102; where Professor Calder reads I read an imperfect N. The photograph shows N, not M.

⁵³ in H. 18 read rather than

⁵⁴ Zingerle first saw that μόγις is needed where Hogarth and I were convinced that the stone has Π⊏ certainly. This exemplifies a class of errors due to climatic conditions: a line on the stone is lengthened by moisture freezing. In this way ГΙ becomes Π.

⁵⁵ Correction kindly sent by Mr. Buckler.

⁵⁶ Some might be taken as natives who gained civitas, but I reject this view: civitas would rarely be conferred in this period and this hardly romanised Region.

⁵⁷ απ' ἑ. π is often written for φ, and vice versa, in Anatolia.

⁵⁸ The imperial police and secret service of the third century (and even late second) were more dreaded than the criminals: they had great opportunity under pretence of exacting contributions for Imperial purposes (κολλητιῶνες).

⁵⁹ Hogarth, in J.H.S., VIII, 1887Google Scholar.

⁶⁰ In 1. 6 read: cattle large or small (Buckler): in

⁶¹ Such is the case now in Greece, or was in 1881.

⁶² One ear would indicate local injury: two ears deafness. See note, p. 287.

⁶³ If Apoll[onios] were the father the phrase would have been δίς. The repetition is rare, but occurs in No. 31, 56.

⁶⁴ The People of Thiounta, , ‘unique … of very small stature and ugly yet not forbidding features, unlike any other people known to me, have apparently preserved unmixed and unchanged the type of some old race’ (C.B. p. 124 f.)Google Scholar. They are not orthodox Moslems and are suspicious of strangers. Compare Gelzer's account of the late survival of pagans in Lydia, Phrygia, etc. in his Pergamon im Mittelalter.

⁶⁵ A θρεπτός was frequently married to a daughter, in which case he seems to have been qualified to carry on the family religion like a son. This practice was common, and was a sort of adoption. If the expenses of his foster-parents were paid he was free.

⁶⁶ The general sense is certain: details and names uncertain.

⁶⁷ is not used in these documents: only

⁶⁸ Example M.A.M.A. I, No. 133, when rightly read.