126 INSCBITTI O N S [ROMAN. them ->* , not an accent), down to the epoch of the emperor Marcus Aurelius. In some very rare instances the doubling of consonants is indicated by a sicilicus, a hook ( ) upon them. The double i indicates, in some examples, from Caesar down to Domitiau, the consonantal j (as in cuiius, eiius). To save space, on coins first and afterwards in inscriptions also, two or three or even more letters were joined, especially at the end of the lines, to a nexus or a ligatura. This system of compendious writing, very rare iu the republican epoch, and slowly extending itself during the 1st century, became rather frequent iu the 2d and 3d, especially in Spain and Africa. There is no constant system in these nexus litterarum, but generally the rule is observed that no substantial element of a single letter is to be counted for twice (thus, e.g., ^ is it or ti, not Tit I). In the republican period, the numbers from one to nine are mostly written in the additive form (I II III Illl V VI VII VIII Vllll), and similarly in combination with X, XX, and so on (XXXX, LXXXX) ; V, for five, seems to be a graphic division of X. The ^ of the Chalcidian alphabet, ^, is the numeral for fifty (afterwards and L, which has originally nothing to do with the letter L); the &, 0, is that for a hundred (replaced early by the initial of the word centum, C) ; the <p, 0, is that for a thousand (afterwards M, the initial of milk), of which <T ,OO are only slight graphic alterations. The multiples of a thousand by 10 are written thus tffo (10,000), ^ (100,000). From Q came, by graphic division, J) (not D the letter) for 500 (with ft 5000, ^ 50,000). A peculiar mark (a^i) appears rarely for 500,000 (Hermes, iii., 1868, p. 467). Numerals are usually dis tinguished from letters in the ancient period, down to the end of the republic, by a stroke drawn through them, as in -H-VIR, duo(m) vir(om) -H S duo semis (sestertius), -Q 500; it was afterwards put above them, as in TlVIR, XVIR, imllVlR, duovir, decemvir, sevir. 1 The direction of the writing is, even in the oldest in scriptions, from left to right ; there exists only one very ancient example of an inscription, found at the lake Fucinus, written in a kind of ySowrpoc^Soi/ arrangement (H. Jordan, Hermes, vol. xv., 1880, p. 5), while in the Sabellic inscriptions similar arrangements are not in frequent. Each word is separated from the other by a sign of interpunction, which is not wanted, therefore, at the end of lines or of the whole text. Exceptions to this rule occur only in the later period (from the 2d century downwards), and sometimes under special conditions, as when abridged words form the end of the line. Here and there even the different syllables of each word are separated by interpunction. The interpunction is formed by a single dot (except in some very ancient inscriptions, such as those of Pisaurum, where, as in Greek and other Italian monu ments, three dots are used), which, according to the technical skill of the different periods in stone-cutting, is in some very ancient inscriptions quadrangular, or similar to an oblique cross ( x ), or oblong (as a bold stroke), but, as a rule, triangular, and never circular. This triangular dot changes, by ornamentation, into a hook (7) or a leaf O); the ivy-leaf-shaped dot is especially frequent in in scriptions from about the 2d century downwards. The dot is always placed at the middle height of the letters, not, as now, at the foot of the line. In large texts of in struments the interpunction is often omitted ; in the later period it is often entirely wanting ; and in short texts, in the disposition of the lines, in the varying sizes of the letters employed, in the division of words at the end of the lines, &,c., certain rules are observed, which cannot be detailed here. In some instances older inscriptions have 1 For other details of numerical notation, fractions, &c., see the manuals of metrology. been cancelled and more recent ones substituted (e.y. } on milestones), especially in the case of the damnatio me morise (in cases of high treason), in consequence of which the names of consuls and emperors are often cancelled ; but in modern times also inscriptions have been delibe rately destroyed or lost ones restored. For understanding the texts of the inscriptions an accu- jate knowledge of the system of abbreviations used in them is necessary. These are almost invariably Utterse singulares: that is to say, the initial letter is employed for the entire word (in all its grammatical forms), or, if one initial, as belonging to more than one word, is not sufficiently clear, the first two or even the first three letters are employed ; rarely more than three. Abbreviations in the true sense of the word (by dropping some letters at the end) are to be found, in the older period, only at the end of lines, and not frequently. In the later period some instances of them have been observed. The litterce singulares, as Valerius Probus taught, are either generally employed (imis generalis) in all classes of written documents (and so in literature also), as, for instance, those of the indi vidual names (the preenomina), the names of days and feasts (hal. for kalfndst), and those of the chief magistrates (cos. for consul) and the like ; or they belong chiefly (but not exclusively) to certain classes of documents, such as those used in juridical acts (I. for lex, h. for hcres, s. d. m. for sine dolo malo, and so on), in sepulchral inscriptions (h. s. e., hie situs est) or in dedicatory inscriptions (v. s. I. m., votum solvit libens merito), &c. 2 It may be observed here that the przenornina are, as a rule, always written in the universally known abbrevia tions (in the few instances where they are written in full, it is a consequence of Greek influence or of peculiar cir cumstances). The gentilicia in -iiis are abridged, in the republican period, in -i (in the nominative, perhaps for -is). In the always abbreviated indications of ancestors or patrons (in the case of slaves and freedmen), as C. /., Gai filius, M. L, Marci libertus (s. for servus is not frequent), the feminine gender is sometimes indicated by inversion of the letters. Thus 0. I. (or lib.) or W (an inverted M) /. designates a mulieris libertus ; 1 and T are used for filia, pupilla. On the tribus and their abbreviations, and on the so-called military tribus (which are names of colonies collocated, for the sake of symmetry, at the place usually occupied, in the nomenclature, by the tribus), and on the other indications of origin used in the designation of in dividuals, the indexes to the above-named works give sufficient information ; on the geographical distribution of the tribus, see Grotef end s Imperium Romamim tributim de- scriptum (Hanover, 1 863). For the abbreviations of official charges, urban and municipal, and, in the imperial period, civil and military (to which, beginning with the 4th cen tury, some Christian designations are to be added), see also the explanations given in the indexes. Among these abbre viations the first instances are to be found of the indication of the plural number by doubling the last letter; thus Augg., Caess., coss., dd. nn. (clomini nostri), are used from the 3d century downwards (see De Rossi s preface to the Inscriptions Christ, urbis Romoe) to distinguish them from Aiiy., Caes., as designating the singular. In the later period, a dot or a stroke over the abridged word, like that upon numerals, here and there indicates the abbreviation. 2 On the system of Roman nomenclature and the abbreviations em ployed in it, see Orelli, cap. viii. (with Wilmauns s Analecta, ii. p. 197), and especially Mommsen in RSmische Forschun gen, vol. i. p. 1 sq., and in Hermes, iii., 1869, p. 70; on the cognomina (but only those occurring in ancient literature), Ellendt (De cognomine et agnomine Romano, Konigsberg, 1853), and on the local cognomina of the Roman patriciate, Mommsen, Rom,. Forsch., ii. p. 290 sq. ; on the nomina gen tilicia, Hiibner (Ephcm. epiyr., ii. p. 25 sq.). The indexes to Orelli, Wilmanns, and the volumes of the Corpus may also be consulted.
