^{page 63 note 1} Archaeologia, xlix. pp. 367–381; liii. pp. 147–154.

^{page 64 note 1} I wish cordially to acknowledge the assistance of the authorities, and especially of the Keeper, of the Department of Greek and Roman Antiquities in the British Museum; and to record my thanks both to the President and Council of the Leeds Philosophical and Literary Society for their kind permission to study and publish these sculptures, and to their Curator, Mr. H. Crowther, for his ready assistance throughout my work, and for kindly taking the photographs of the fragments at Leeds which are illustrated in this article.

^{page 64 note 2} Archaeologia, xlix. p. 381.

^{page 65 note 1} Exceptional vehicles must be distinguished from chariots in this respect. E.g. Poseidon and Amphitrite appear seated in a low car drawn by a Triton in an Augustan relief, cf. Strong, Roman Sculpture, p. 35, and Pl. V.

^{page 65 note 2} Archaeologia, xlix. Pl. 27, 1.

^{page 65 note 3} For a chariot-wheel with plain baluster-spokes, not of course floreated, cf. that on the Bosco Reale cup representing the triumph of Tiberius (Strong, Roman Sculpture, Pl. XXIV. 1).

^{page 65 note 4} For the motive of closed palm-leaves cf. the candelabrum figured in Daremberg and Saglio, s.v., Fig. 1092.

^{page 65 note 5} These will be published and illustrated next year in the second part of this article. The dimensions are: (a) (larger fragment), h. ·40 m., circ. at bottom, ·38, at ca. ·25 above this, ·345; (b) h. ·18, circ. at bottom ·25. Assuming a regular decrease of ·035 in ·25 m., rather less than one metre separated the bottom of a from that of b, and thus slightly over half a metre is missing from between the fragments. Thus the minimum height of the shaft was ·40 + ca. ·52 + ·18 ( = ca. 1·10); it may have been taller still.

^{page 66 note 1} Archaeologia, xlix. p. 379, ‘a parallelogram of massive construction. … The walls running in the direction of north and south are nearly twelve metres long, the cross walls about six metres.’

^{page 66 note 2} See below, pp. 90 f.

^{page 66 note 3} Below, pp. 86 f.

^{page 66 note 4} Where it is not stated to the contrary these are in the Museum at Leeds. All the pieces except the six torsi and the four large fragments of horses are reproduced on a scale of one-sixth.

^{page 66 note 5} Nearly all the fragments here described are of a fine-grained white marble, which varies between almost pure white and a brownish shade according to circumstances. Mr. A. J. B. Wace confirms my opinion that this is an Italian, not a Greek, marble.

^{page 66 note 6} For the type of lorica see below, p. 89.

^{page 68 note 1} See below, p. 88, for discussion of details of the costume.

^{page 69 note 1} Entered in the inventory of the Museum among fragments of horses' legs.

^{page 70 note 1} Found in two pieces separately inventoried, and joined by me. The edge of this fragment is dressed smooth where it joined the upper part of the thigh.

^{page 71 note 1} For this type of boot cf. the ‘Alexander’ statuette, referred to in the next note. It seems to be that called caliga; for other examples cf. D. and S. s.v., Fig. 1033, from the column of Trajan, and the standing figure of Marcus Aurelius in the British Museum, B.M. Sculpture, iii. 1906 Google Scholar. It is clearly of Greek origin, and differs in no way from those worn by two figures in the group of Daochus at Delphi, namely Sisyphus II. and Acnonius, Fouilles de Delphes, Pl. LXV., LXVI., cf. A.J.A. xiii. (1909), p. 450 Google Scholar, Fig. 3. A lower shoe of the same type is worn also by the seated figure (Pluto ?) on the Hermes column from Ephesus in the British Museum, and the seated Asklepios on a relief from Epidaurus (see Lethaby, , J.H.S. xxxiii. (1913), p. 93 and Fig. 4Google Scholar). Thus it is at least as old as the 4th century B.C.

^{page 71 note 2} Brunn, Denkmäler der Gr. und Röm. Sculptur, No. 355 b.

^{page 71 note 3} Found in two pieces and mended by me.

^{page 75 note 1} As mentioned by Savile, Lord, op. cit. xlix. p. 381 Google Scholar; these two fragments were found at different portions of the site.

^{page 76 note 1} It seems unlikely that the muzzle No. 29 could have been made in a separate piece, for there is no trace of a dowel to hold it to the upper part of the head, as in No. 28; moreover the presence of a hole for the bridle in the left cheek of both fragments proves that they belong to separate heads. The difference in marble and workmanship precludes No. 30 from being combined with No. 28. Thus we must admit the existence of seven horses' heads, as was recognised by the discoverer, op. cit. liii. p. 150.

^{page 77 note 1} Found in two pieces and mended by me.

^{page 79 note 1} Joined by me; the mending has not quite brought the hind surface of the two pieces into a straight line. These perhaps belong to the same leg as No. 39.

^{page 79 note 2} Probably indicating an ancient breakage.

^{page 81 note 1} Probably Nos. 53 and 54 formed the 1. hind-leg of the horse of which No. 50 is the r.

^{page 83 note 1} The flat plane in which this fragment is worked, its thinness, and its smooth back suggest that it might be rather from a relief than a figure in the round; but the treatment of the muscles resembles so closely that of the other horses that it seems safer to connect it with the group, as did Lord Savile and MrSmith, A. H. in B.M. Sculpture, vol. iii Google Scholar.

^{page 85 note 1} The attribution of this fragment to the group is, to say the least, doubtful.

^{page 86 note 1} Found in two pieces and now mended.

^{page 86 note 2} For other instances of horses' tails in ancient marble sculpture reaching to the base on which they stand cf. the equestrian statue of M. Nonius Balbus from Herculaneum at Naples, and the horses on Monte Cavallo at Rome, Furtwängler and Urlichs, , Greek and Roman Sculpture (Eng. Trans. 1914), pp. 44 f., Figs. 13, 14Google Scholar.

^{page 86 note 3} Op. cit. xlix. p. 374.

^{page 86 note 4} Op. cit. p. 375.

^{page 86 note 5} Fragments 62–64 leave no doubt on this point.

^{page 87 note 1} B.M. Sculpt, iii. pp. 103 f.

^{page 87 note 2} Cf. Röm. Mitt. xii. (1897), p. 179 Google Scholar.

^{page 87 note 3} It is just possible that No. 28 should be rejected from the group; but there is no reason for limiting the number to six, as one (or more) of the riders' figures may have completely disappeared.

^{page 87 note 4} On Roman reliefs and coins they are more frequently represented as going to r.; the shoulder further from the spectator is usually more raised and advanced. Greek art allowed more variety on such a point (e.g., contrast the legs of the horses on Syracusan Dekadrachms by Kimon and Euainetos.)

^{page 88 note 1} E.g., on the figure of M. Aurelius on the relief, in the Palazzo dei Conservatori in Rome, showing the Emperor pardoning captives.

^{page 88 note 2} See Daremberg and Saglio, s.vv. Lorica, §§ 15, 26, and Cingulum, esp. Fig. 1501.

^{page 88 note 3} To either of these might be attributed the fragment No. 15, which is apparently of a rider's r. leg pressed against his horse's flank as he turned to the right in his saddle.

^{page 89 note 1} P. 71, note 2.

^{page 90 note 1} The best examples of this type occur in coins, where it is presumably adapted from a statue-type. Cf. the reverses of gold and ‘first brass’ coins of Trajan, Cohen, Medailles ², Nos. 501, 503. It is common on copper coins of 3rd century Emperors, and was adopted occasionally by Greek cities in the provinces; e.g. Tomi, where it appears on a bronze coin of Gordian III., Pick, Antike Münzen Nordgriechenlands, Pl. VII. 19. Other examples of equestrian statues with human and semi-human figures under the horses' fore-legs are collected in Reinach, , Rep. Sculpt. ii., pp. 528 Google Scholar, 5, 6, 529, 1–5, 533, 4; iii. p. 270, 2; iv. p. 334, 2.

^{page 90 note 2} Cf. the use of trophies of armour to support the standing figures of the Dioscuri, on Monte Cavallo (above, p. 86, note 2), where the horses are supported with plain marble pedestals behind the fore-legs. For tree-trunks used for this purpose see Amelung, , Sculp. des Vatican. Mus. vol. ii.Google Scholar Pls. XXXVII, No. 166; XXXVIII, Nos. 175, 178, which are all very much restored.

^{page 91 note 1} The enthusiasm of Lord Savile's friend and correspondent, in Archaeologia, liii. p. 151, is not shared by the present writer, after repeated handling of all the fragments.

^{page 91 note 2} Op. cit. liii. p. 149 f. I did not see, until this article was in print, a valuable paper on the topography of Civita Lavinia by Colburn, G. E., in A.J.A. xviii. (1914), pp. 185 ffGoogle Scholar. It should be noted that he does not accept this site as a Royal Villa (pp. 197 f.); that he considers these sculptures to be of Parian marble (merely following Lord Savile), and is apparently not aware of the existence of the collection at Leeds.

^{page 91 note 3} There is no difficulty about believing that the group may have represented members of the Imperial family and mounted retainers. But it is to be hoped that the portrait-heads were of better style than the torsi, which may well have been the case.