^{page 147 note 1} While I accept from Herodotus the fact of Dorian recognition of Orestes, I do not believe that his account is anything more than a popular legend to explain an existing cultus, or that the date about 560 which he assigns can be taken as historical. The ignorance of the forging of iron implies an earlier origin even for the legend.

^{page 147 note 2} To quote these words as a proof that Cleomenes was really of a non-Dorian family, as has been done, seems to me a misunderstanding of the nature of Greek legend.

^{page 148 note 1} Their fig. 117 is in some respects more successful than my fig. 9; but there are two faults in it. (1) It is the right warrior, not the left, which is complete: my fig. 7 represents the relief from the opposite view, viz. from the interior. (2) The rows indicating the hair of the Gorgon-like figure are not visible in a front view, but only in a side view. They are indicated on the edge of the relief: the head is indicated as a flat surface and the edges are cut sharp and square down to the background. I have omitted to mention this detail in my description. I could detect no attempt to indicate eyes. But the large drawing gives a far better idea of the relief than my tiny figure 9.

^{page 148 note 2} Fig. 128 (cp. p. 105, n. 1) is due to Mr. Blunt, not to me, while fig. 90, which is attributed to Mr. Blunt, is due to me, and differs from the drawing by Mr. Blunt, which is among the papers of the Society.

^{page 149 note 1} The inaccuracy is very slight on the right side, but more serious on the left side.

^{page 149 note 2} So puzzling are these patterns that, although the error relates to a point which has particularly interested me, I had looked cursorily many times at the drawing without observing the error. One can never be sure of having understood the pattern without drawing it with one's own hand.

^{page 149 note 3} Mr. Blunt intimated in the Journal his readiness to supply copies of this photograph at a very small price.

^{page 150 note 1} The incised parts are thinner than the raised parts at Ibriz; M. Guillaume shows the raised parts thinner than the incised parts on the Midas-Tomb.

^{page 150 note 2} Journal of Hellenic Studies, 1882, pp. 26–7, and note on p. 28.

^{page 151 note 1} Perrot fig. 61, 62.

^{page 151 note 2} It would perhaps be move accurate to say that the lines of the fundamental pattern are arranged obliquely instead of horizontally and vertically. In that way also trial will show that if the four lozenges coincide with the lines of the fundamental pattern, the squares between the four lozenges have not their angles on the fundamental lines. The fundamental lines are, of course, purely imaginary, and I speak of them only in order to bring out as clearly as possible the actual character of the Phrygian pattern.

^{page 152 note 1} Part I. p. 371; J. H. S. 1884, p. 254ff. M. Perrot somewhat strangely has not noticed (p. 102) that there is a door on this monument.

^{page 153 note 1} The two pediments ave even more alike than can be gathered from pl. xxi A. The supporting column of the pediment in each is of the same type, with a rectangular capital and base, the base smaller than the capital.

^{page 154 note 1} Formerly I thought that Greek art exercised great influence in this period, but I have been taught better by Professor G. Hirschfeld. I see much Persian influence (differing in this from Hirschfeld) and a little Greek, the latter very late.

^{page 155 note 1} I have not access to Texier's large work and have to content myself with the drawing published in his small work, Asic Mineure.

^{page 155 note 2} He gives the right spiral distinctly, while the left which is uncertain must be understood to resemble the right.

^{page 156 note 1} La première hypothèse qui se présente à l'esprit, p. 89. I should rather say that the most obvious reflexion, which rises in everyone's mind on first seeing the monument, is that it cannot be a tomb, as there is no apparent place for a grave.

^{page 156 note 2} The Phrygian mysteries, as to whose rites we are well informed, are a presentation in gross symbolism (according to the primitive social circumstances and the elementary ideas of nature which existed at the time) of this religious idea. The ideas entertained by the Greeks about Cybele are in the main Greek and not Phrygian, and should have no weight attached to them.

^{page 156 note 3} On p. 102, n. 1, he gives it (sic!) On p. 14 he infers from this misquotation that ‘Midas se confondait avec un de ces dieux dont le culte resta populaire,’ &c.

^{page 157 note 1} compare C.I.G. 4033. 4034 &c. I published the inscription of Anaboura in Mittheilungen, Athen, 1883, p. 71. It has since been published by Prof. Sterrett, J. R. S. in his ‘Preliminary Report,’ p. 14Google Scholar, without observing the previous publication.

^{page 158 note 1} It deserves note that all the persons mentioned on the monument have a double name, and that the double name is characteristic of Phrygia in the later inscriptions written in the Greek language—Ates Arkiaevais, Midas Lavaltas the King, Baba Memevais.

^{page 159 note 1} Monument commemoratif, p. 102.

^{page 160 note 1} I have on p. 376 mentioned the tendency of this rock to split in vertical surfaces.

^{page 161 note 1} This discovery is no matter of conjecture. I think that any one who examines the rock will come to the conclusion there is no other way of accounting for the loss of the epsilon, except through actual cleavage of the rock. Mere mouldering of the surface does not account for the loss. I intended to reexamine the place in 1887 along with Mr. Hogarth and get his testimony, but the Circassians, who have recently built a village beside the Midas-Tomb, have constructed a store-room in front of the chamber.

^{page 162 note 1} I know what almost insurmountable difficulties there are to prevent a good photograph being obtained, on account of the position and surroundings of the monument.

^{page 162 note 2} Only three letters extend beyond the ornament in the line below the pediment. Texier made his letters too broad in proportion to their height (all Phrygian letters are tall and thin) and thus makes eight extend beyond the ornament. He could not get the inscription from the photograph, for, as I explain on p. 382, some of the letters are restored. The third, fourth and and fifth words are so mutilated that they long baffled all copyists, including myself in 1881 and Sterrett and myself in 1883. In 1884 I made them out with a good glass, and Hogarth entirely agreed with my copy in 1887. I have restored the letters completely, but there remain only the tops of the ten middle letters.

^{page 162 note 3} These lozenges are, I think, true squares whose diagonals are at right angles to the sides of the pediment. This is probably true in all monuments of this class, though the point is difficult to determine on a distant view.

^{page 162 note 4} A swallow's nest perched between the two horns of the acroteiion is represented and exaggerated in M. Perrot's drawing, and in the sketch by Mr. Blunt, which I showed to M. Perrot and which is attributed to me.

^{page 163 note 1} M. Perrot aud I agree as to the number of squares in the horizontal band of ornament below the inscription. Sir C. Wilson thinks we have one too many, and I think he is right; yet it is hard to believe that M. Perrot, with a photograph before him (which I have not seen), could be wrong on such a point.

^{page 163 note 2} The words aez and atanizen. kurzanezon. ta. When I suppressed the second drawing which I originally intended to give, I added these words at the side of the first, thus making it inaccurate; but I wished to give the inscription complete. From the point of view of fig. 13 these symbols at the side cannot be seen.

^{page 163 note 3} The ‘croquis de M. Ramsay’ used to correct Texier is really Mr. Blunt's drawing. Mr. Blunt was successful with this and with the Midas-Tomb: he is not represented in the Journal by his best work.

^{page 164 note 1} These points are of course of no practical importance, as they do not affect the ornamentation. I merely mention them for the sake of completeness.

^{page 164 note 1} See the map, which M. Perrot has given fig. 47. The monument which is here given as fig. 18 is near the one which is there numbered 3.

^{page 165 note 3} Redrawn as usual, without the slightest alteration in character, by Mr. McCann.

^{page 166 note 1} The words ‘Tomb with relief of hunt’ refer to the same monument, and I wished them to be erased from the proof of the map.

^{page 166 note 2} In the large drawing from which Fig. 11 is reduced, the various remains of fortification were shown in different colours. One who wishes to understand the arrangement of the fortifications must colour the remains in Fig. 11 in order to see them readily. The plan has been so much reduced in scale that it has lost all clearness.

^{page 167 note 1} The word OUTWORK is by mistake printed a little too high in Fig. 12. It is placed almost outside of the probable line of fortification and on the dromos.

^{page 167 note 2} The illustrations have now appeared in the Athenische Mittheilungen, 1889, p. 170ff. My reasons for holding these reliefs to be ancient, are (1) they are in all probability made along with the dromos, (2) the curve of the back, which seems of late style to M. Perrot, appears in figures which are unmistakably ancient (see above, Fig. 18, and Mittheil., Fig. 4).

^{page 167 note 3} A step too many is represented in front of the altar in the illustration. The drawing from which it was taken was done by Mrs. Ramsay in 1881, merely to assist her memory without any thought of publication. The task of preparing drawings for publication belonged to Mr. Blunt, who however had in truth not time enough to do himself justice.

^{page 169 note 1} This inscription differs only in one word and two letters from the one on the right side of the Midas-Tomb. I advance a suggestion about its interpretation at the end of this paper.

^{page 171 note 1} I may here once for all acknowledge the skill with which Mr. McCann has from my measurements reproduced the form of these altars in perspective. To make drawings of objects which he had never seen was a very difficult task, and has been performed very skilfully.

^{page 171 note 2} If my theory that the Midas monument is a grave is untrue, I see no other possibility except to ascribe a similar character to it, to Perrot's Fig. 59, and to the monument given in my Historical Relations, Plate III., Fig. 10. These are all carved under the city walls. As I have stated above, M. Perrot's theory that they are commemorative cenotaphs suits none of the facts.

^{page 174 note 1} Stewart's copy is barely recognisable as the same inscription, though in general his copies of the Phrygian inscriptions are better than those of Texier, Mordtmann, or even Leake.

^{page 177 note 1} Fire-places of the very same type are in use at the present day.

^{page 177 note 2} M. Perrot, on p. 77, attributes to me an opinion, which I never for a moment held, that these rooms were bed-room, dressing-room, and bath-room. A sitting-room is a necessary part of a harem, and a large chamber with a fireplace can never have been used for a bed-room. He also, on p. 76, makes the larger northern room of the ἀνδρωνιτις a Christian chapel; it is the smaller middle chamber that has been used for that purpose.

^{page 178 note 1} ‘Paphlagonische Felsengräber’ in Berl. Akad. Abhandl., 1885. I am glad to agree with almost everything that Hirschfeld says about the relations between Greek and Phrygian art, though I have been forced to dissent from some of his opinions on Syro-Cappadocian art (see Athenische Mittheilungen, 1889, p. 171 ff).

^{page 178 note 1} The journey of 1884, in which Mr. A. H. Smith co-operated with me till his health failed, was the only one in which I have had anything like proper equipment for accurate work; but the Asia Minor Exploration Fund could not stand another expedition on such a scale.

^{page 179 note 1} Stewart represents one gorgoneion clearly, and I felt no hesitation in identifying the ornament as a gorgoneion; but MM. Perrot and Guillaume were not so certain about it. But even M. Perrot admits that a number of smaller gorgoneia exist (Fig. 87).

^{page 179 note 2} I showed to him the very same illustrations (drawn in Oxford in 1885), which have been reproduced as Figs. 29–33.

^{page 179 note 3} A method of construction similar to that of the ‘Wall of Romulus’ on the Palatine.

^{page 181 note 1} Such a stair and spring may bo seen still perfect in the Acropolis of Amasis on the top of a conical rock 1,200 feet above the level of the Iris and its narrow plain. I should have examined the stair at Yapuldak more thoroughly if I had carried out my intention to return and map the Acropolis completely.

^{page 182 note 1} The roof of the later chambers is vaulted.

^{page 182 note 2} Graves of the arcosolium type occur by the hundred in the Phrygian Necropoleis.

^{page 185 note 1} The gorgoneion in fig. 30 is sculptured in flat relief, so that the features are almost on one plane, and the edges round the face are cut square down to the wall of the chamber. The gorgoneion is represented as looking down into the chamber, the upper part of the head projecting several inches further from the background than the lower part. This character distinguishes it from Greek work.

^{page 185 note 2} Viz., Baba Memefais Proitafos kwizanafezos sikeneman egaes. The engraver of the other text has omitted two letters, a in Baba and z in kwizanafezos. The omission is probably accidental.

^{page 186 note 1} Deecke, (Lyk. Studien p. 318 in Bezz. Beitr. vol. xiiGoogle Scholar) also makes cdaes equivalent to ἔθηκε. Fick has shown that the Phrygian glosses prove the aspirates to appear in Phrygian as sonants

^{page 187 note 1} I have now unconsciously adopted an interpretation of which was advanced some years ago by Professor Sayce.

^{page 187 note 2} De monumentis ad Odysseam pertinentibus, p. 5. Dümmler's explanation, (see Berl. Philolog. Wochenschr., 1888, p. 17), seems to me inadmissible. Anything can be made out of an inscription if we may insert letters ad lib.

^{page 188 note 1} See Bergk, in Zft. f. Numismatik, 1884, p. 333Google Scholar. He argues that Aspendos, which is called an Argive colony, was founded by Achaean Argives, who had gone to Cyme. Selge, an Amyclaean colony, might be explained in a similar way (Dionys. Perieg. 860 and Eustath. ad loc).

^{page 188 note 2} The form is rather blurred, but there can be little doubt about it.

^{page 188 note 3} It may however be safely asserted that ψ in Phrygian is not the Ionic psi. Phrygian used ΚΣ, not xi, and cannot have adopted psi before xi.

^{page 188 note 4} Eine vorgriech. Inschrift aus Lemnos, p. 17.

^{page 189 note 1} The analogy with Phrygian so struck me at the first glance, that I immediately concluded they were in the Phrygian language, till examination showed that they were certainly in a different language.

^{page 189 note 2} After this paragraph was in type Professor Hirsehfeld's article in Rhein. Mus. 1889, p. 461, appeared. He considers, rightly as I think, that the so-called Ionic alphabet is simply the alphabet of Miletos, which gradually was adopted, first by the other Ionic cities, and finally by the whole of Greece.