ravine or gully. The typical species, I guanodon mantelli, measures 5 to 6 metres in length, while I. bernissartensis (see ng.) attains a length of 8 to 10 metres. They are found both at Bernissart and in the south of England, while other species are also known from Sussex. Nearly complete skeletons of allied reptiles have been discovered in the lurassic and Cretaceous rocks of North America.
REFERENCES.-G. A. Mantell, Petrifactions and their Teaching (London, 1851); L. Dollo, papers in Bull. Mus. Roy. d'Hist. Nat. Belg., vols. i.-iii. (1882-1884). (A. S. WO.)
IGUVIUM (mod. Gubbio, q.v.), a town of Umbria, situated among the mountains, about 23 m. N.N.E. of Perusia and connected with it by a by-road, which joined the Via Flaminia near the temple of Jupiter Appenninus, at the modern Scheggia. It appears to have been an important place in pre-Roman times, both from its coins and from the celebrated tabiilae Igzwirzae (see below).
We find it in possession of a treaty with Rome, similar to that of the Camertes Umbri; and in 167 B.C. it was used as a place of safe custody for the Illyrian King Gentius and his sons (Livy xlv. 43). After the Social War, in which it took no part, it received Roman citizenship. At that epoch it must have received full citizen rights since it was included in the tribus Clustumina (C.I.L. xi. e.g. 5838). In 49 B.C. it was occupied by Minucius Thermus on behalf of Pompey, but he abandoned the town. Under the empire we hear almost nothing of it. Silius Italicus mentions it as subject to fogs. A bishop of Iguvium is mentioned as early as A.D. 413. It was taken and destroyed by the Goths in 552, but rebuilt with the help of Narses. The Umbrian town had three gates only, and probably lay on the steep mountain side as the present town does, while the Roman city lay in the lower ground. Here is the theatre, which, as an inscription records, was restored by Cn. Satrius Rufus in the time of Augustus. The diameter of the orchestra is 76% ft. and of the whole 230 ft., so that it is a building of considerable size; the stage is well preserved and so are parts of the external arcades of the auditorium. Not far off are ruins probably of ancient baths, and the concrete core of a large tomb with a vaulted chamber within. (T.As.)
Of Latin inscriptions (C.I.L. xi. 5803-5926) found at Iguvium two or three are of Augustan date, but none seem to be earlier. A Latin inscription of Iguvium (C.I.L. xi. 5824) mentions a priest whose functions are characteristic oi the place “ L. Veturius Rufio avispex extispecus, sacerdos publicus et privatus.” The ancient town is chiefly celebrated for the famous Iguviiic (less correctly Eilgubine) Tables, which were discovered there in 1444, bought by the municipality in 1456, and are still preserved in the town hall. A Dominican, Leandro Alberti (Descrizione d'Ilalia, 1550), states that they were originally nine in number, and an independent authority, Antonio Concioli (Stalula civilalis Ezrgubii, 1673), states that two of the nine were taken to Venice in 1540 and never reappeared. The existing seven were first published in a careful but largely mistaken transcript by Buonarotti in 1724, as an appendix to Dempster's De Etruria Regal#
The first real advance towards their interpretation was made by Otfried Muller (Die Elrusker, 1828), who pointed out that though their alphabet was akin to the Etruscan their language was Italic. Lepsius, in his essay De labulis Eugubinis (1833), finally determined the value of the Umbrian signs and the received order of the Tables, pointing out that those in Latin alphabet were the latest. He subsequently published what may be called the cditia prihceps in 1841. The first edition, with a full commentary based on scientific principles, was that of Aufrecht and Kirchhoff in 1849-1851, and on this all subsequent interpretations are based (Bréal, Paris, 1875; Biicheler, Umbrica, Bonn, 1883, a reprint and enlargement of articles in Fleckeisen's Jahrbuch, 1875, pp. 127 and 313). The text is everywhere perfectly legible, and is excellently represented in photographs by the marquis Ranghiasci-Brancaleone, published with Bréal's edition.
- A portion of this article is taken by permission from R. S.
Conway's Italic Dialeds (Camb. Univ. Press, 1897). Language.-The dialect in which this ancient set of liturgies is written is usually known as Umbrian, as it is the only monument we ossess of any length of the tongue spoken in the Umbrian distri t before it waslatinized (see UMBRIA). The name, however, is certainly too wide, since an inscription from Tuder of, probably, the 3rd century B.C. (R. S. Conway, The Italic Dialecls, 352) shows a final -s and a medial -d-, both apparently preserved from the changes which befell these sounds, as we shall see, in the dialect of Iguvium. On the other hand, inscriptions of Fulginia and Assisium (ibid. 354-355) agree very well, so far as they go, with lguvine. It is especially necessary to make clear that the language known as Umbrian is that of a certain limited area, which cannot yet be shown to have extended very far beyond the eastern half of the Tiber valley (from Interamna Nahartium to Urvinum Mataurense), because the term is often used by archaeologists with a far wider connotation to include all the Italic, pre-Etruscan inhabitants of upper Italy; Professor Ridgeway, for instance, in his Early Age of Greece, frequently speaks of the “ Umbrians " as the race to which belonged the Villanova culture of the Early Iron age. It is now one of the most urgent problems in the history of Italy to determine the actual historical relation (see further ROME: History, acl. init.) between the 'Oatipoi of, say, Herodotus and the language of lguvium, of which we may now offer some description, using the term Umbrian strictly in this sense.
Under the headings LATIN LANGUAGE and OscA LINGUA there have been collected (1) the points which separate all the Italic languages from their nearest conveners, and (2) those which separate Osco-Umbrian from Latin. We have now to notice (3) the points in which Umbrian has diverged from Oscan. The first of them antedates by six or seven centuries the similar change in the Romance languages (see Romance LANGUAGES).
(1) The palatalization of k and g before a following i or e, or consonant i as in ticil (i.e. digit)=Lat. decel; muieio past part. passive (pronounced as though the i were an English or French j) beside Umb. imperative mugalu, Lat. mugire.
(2) The loss of final-d, e.g. in the abl. sing. fem. Umb. t6ld=Osc. loutrid.
(3) The change of d between vowels to a sound akin to r, written by a special symbol <1 (gl) in Umbrian alphabetiand by RS in Latin alphabet, e.g. teda in U mbrian alphabet=dirsa in Latin alphabet (see below), “ let him give, ” exactly equivalent to Paelignian dida (see PAELIGNI).
(4) The change of -5- to -r- between vowels as in erom, “ esse " =Osc.'ezum, and the gen. plur. fem. ending in -aru=Lat. -arum, Osc. -azum.
To this there appear a long string of exceptions, e.g. as a = Lat. ara. These are generally regarded as mere archaisms, and unfortunately the majority of them are in wo1'ds of whose origin and meaning very little is known, so that (for all we can tell) in many the -5- may represent -ss- or -ps- as in osatu =Lat. operate, cf. Osc. opsaom. (5) The change of final -ns to -f as in the acc. plur. masc. 'z/itluf-Lat. vitulés.
(6) In the latest stage of the dialect (see below) the' change of final -s to -r, as in abl. plur. arver, arviis, i.e. “ arvorum frugibus." (7) The decay of all diphthongs; ai, oi, ei all become a monophthongs variously written e andi (rarely ei), as in the dat. sing. fem. tote, “ civitati ”; dat. sing. masc. pnple, “ populo ”; loc. sing. masc. ohse (lrom *om(e)sei), “ in umero." So au, eu, oil all become 6, as in ole=Osc. anti, Lat. aut.
(8) The change of initial l to 1/, as in vulu =Lat. lar/ilo. Owing to the peculiar character of the Tables no grammatical statement about Umbrian is free from difficulty; and these bare outlines of its phonology must be supplemented by reference to the lucid discussion in C. D. Buck's Oscan and Umbrian Grammar (Boston, 1904), or to the earlier and admirably complete Oskischambrische Grammaiik of R. von'Planta (Strassburg, 1892-1897). Some of the most important questions are discussed by R. S. Conway in The Italic Dialects, vol. ii. p. 495 seq.
Save for the consequences of these phonetic changes, Umbrian morphology and syntax exhibit no divergence from Oscan that need be mentioned here, save perhaps two peculiar perfect-formations with -l- and -ngi-; as in ampeluxt, fut. perf. “ impenderit, " combijiangiust, “ nuntiaverit ” (or the like). Full accounts of the accidence and syntax, so far as it is represented in the inscriptions, will be found in the grammars of Buck and von Planta already mentioned, and in the second volume of Conway, op. cii. Chronology. (l.) The Relative Dates of the Tables.-At least four periods in the history of the dialect can be distinguished in the records we have left to us, by the help of the successive changes (a) in alphabet and (b) in language, which the Tables exhibit. Of these only the outstanding features can be mentioned here; for a fuller discussion the reader must be referred to The I talicDialects, pp. 400 sq . (a) Changes in Alphabet.-Observe first that Tables ., ll., Ill. and IV., and the first two inscriptions of V. are in Umbrian character; the Latin alphabet is used in the Claverniur paragraph (V. iii.), and the whole of VI. (a and b) and VII. (a and b). What we may call the normal Umbrian alphabet (in which e.g. Table l. a is written) consists of the following signs, the writing being always from right to left: Fl a, 8 b, 'I gi (aa. a. sound akin to r derived from d). 30, 711, 4 2, 0 h, I 1,51 /¢ and g, |l, H-1 m, H n,