THE UMBRIAN LANGUAGE
In the neighbourhood of Ugubio, at the foot of the Apennines (the ancient Iguvium), were discovered, in A.D. 1444, seven tables, commonly called the Eugubine Tables. They were in good preservation, and contained prayers and rules for religious ceremonies. Some of them were engraved in the Etruscan or Umbrian characters, others in Latin letters. Lepsius has determined, from philological considerations, that the date of them must be as early as from A.U.C. 400, and that the Latin were engraved about two centuries later. A comparison of the two shows, in the Umbrian character, the letter s standing in the place occupied by r in the Latin, and k in the place of g, because the Etruscan alphabet, with which the Umbrian is the same, did not contain the medial letters B, G, D. An analogous substitute is seen in the transition from the old to the more modern Latin. The names Furius and Caius, for example, were originally written Fusius and Gaius. H is also introduced between two vowels, as stahito for stato, in the same way that in Latin alieneus is derived from aes. It also appears that the termination of the masculine singular was o: thus, orto = ortus; whilst that of the plural was or; e.g., subator = subacti; screliitor = scripti. This mode of inflexion illustrates the form amaminor for amamini, which was itself a participle used for amamini estis, an idiom analogous to the Greek τετυμμένοι εἰσί.
The following extract, with the translation by Donaldson, together with a few words which present the greatest resemblance to the Latin, will suffice to give a general notion of the relation which the Umbrian bears to it:
Teio subokau suboko, Dei Grabovi, okriper Fisiu, tota-per Jiovina, erer nomne-per, erar nomne-per; fos sei, paker sei, okre Fisei, Tote Jiovine, erer nomne, erar nomne.
Te invocavi invoco, Jupiter Grabovi, pro monte Fisio, pro urbe Iguvina, pro illius nomine, pro hujus nomine; bonus sis, propitius sis, monti Fisio, urbi Iguvinae, illius nomine, hujus nomine.
Ancient Form Latin Form Meaning
Alfu albus white
Asa ara altar
Aveio avis bird
Buf boves oxen
Fetu fetus son
Nep nepos son
Nome nomen name
Parfa parra owl
Pefa probus pious
Periklum periculum prayer (dim.)
Poplo populus people
Punis pontes bridge
Tefla tabula table
Tupu duplex double
Tripler triplum triple
Tota (analogous to) totus a city (a whole or collection)
Vas fas law
Vinu vinum wine
Uve ovis sheep
Vitlu vitulus calf
THE OSCAN LANGUAGE.
The remains which have come down to us of this language belong, in fact, to a composite idiom made up of the Sabine and Oscan. Although its literature has entirely perished, inscriptions fortunately still survive; but as they must have been engraved long subsequently to the settlement of the Sabellians in Southern Italy, the language in which they are written must necessarily be compounded of those spoken both by the conquerors and the conquered. Although Livy makes mention of an Oscan dramatic literature, for he tells us that the “Fabulae Atellanae” of the Oscans were introduced when a pestilence raged at Rome, together with other theatrical entertainments, he only speaks of the Oscan language in one passage. This, however, is an important one, because it proves that Oscan was the vernacular tongue of the Samnites at that period. He relates that Volumnius sent spies into the Samnite camp who understood Oscan: “Gnaros Oscae linguae exploratum quid agatur mittit.” It is clear that the reason why the Oscan language prevailed amongst this people is, that the dominant orders in Samnium were Sabines. But there is evidence of the existence of Oscan in Italy at a still later period. Niebuhr asserts that in the Social War the Marsi spoke Oscan, although in writing they used the Latin characters. Some denarii still exist struck by the confederate Italian Government established in that war at Corfinium, on which the word Italia is inscribed, whilst others bear the word Viteliu. The latter is the old Oscan orthography, the former the Latin. One class of these coins, therefore, was struck for the use of the Sabine, the other of the Marsian allies. It is said also that Oscan was spoken even after the establishment of the empire.
The principal monument of the Sabello-Oscan is a brass plate which was discovered A.D. 1793. As the word Bansae occurs in the 23rd line of the inscription, it has been supposed to refer to the town of Bantia, which was situated not far from the spot where the tablet was found, and it is therefore called the Bantine Table. In consequence of the perfect state of the central portion, much of this inscription has been interpreted with tolerable certainty and correctness. The affinity may be traced between most of the words and their corresponding Latin; and it is perfectly clear that the variations from the Latin follow certain definite rules, and that the grammatical inflexions were the same as in the oldest Latin. A copy of the Table may be found in the collection of Orellius, and also in Donaldson’s Varronianus. The following are a few specimens of words in which a resemblance to the Latin will be readily recognized, and also, in some instances, the relation of the Oscan to the other ancient languages of Italy:—
Licitud liceto
Aultam voluntatem
Maiimas Maximus
Sannais Samnis
Censetis Census
Pis quis
Comenei Comitia
Pomperias Pompeius
Volotud voluntate
Hipid habet
Prúfatted probavit
Pruhipust prohibuit
Censazet censuit
Vensamús, &c. venerunt, &c.
Fust fuit
Poíad posuit
Fuid fuit
Tarfinius Tarquinius
Auti auctor
To these other well-known words may be added, which all philologers allow to be originally Oscan, but which have become incorporated with the Latin—such as, for example, Brutus, Cassius, Catus, Foedus, Idus, Poena, Tiberis, and names of deities, such as Fides, Terminus, Vertumnus, Fors, Flora, Lares, Mameres, Quirinus, &c.
THE ETRUSCAN LANGUAGE.
The difficulty and obscurity in which the Etruscan language is involved are owing to the nature of the inscriptions and monuments which have been discovered. Those records, to which reference has already been made when speaking of the Umbrian and Sabello-Oscan, were of a ceremonial or legal character; they therefore contained connected phrases and sentences, varied modes of thought and expression. Monuments such as the Eugubine or Bantine Tables contribute not a little towards a knowledge of the language, and still more to a knowledge of their structure and analogies. This, however, is not the case with the Etruscan monuments which have been hitherto discovered. They are, indeed, numerous, but they exhibit little variety. They are sepulchral records of a complimentary kind, or titles inscribed on statues and votive offerings. Hence the same brief phraseology continually recurs, and the principal portions of the inscriptions are occupied by proper names.
The most important because the largest Etruscan record which has been hitherto discovered, is one which was found near Perugia, A.D. 1822. This inscription contains one hundred and thirty-one words and abbreviations of words, and of these no fewer than thirty-eight are proper names. Of the rest, a vast number are either frequently repeated, or are etymologically connected. These have not proved sufficient to enable any philologer (although many have attempted it) to give a satisfactory and trustworthy explanation of its contents. A comparison of the Perugian with the Eugubine inscription shows the existence of similarity between some of the words found in both of them; and this is exactly what we should a priori expect to result from the theory of the Etruscan being a compound of the Pelasgian and Umbrian. In the Perugian inscription, words which resemble the Umbrian forms are more numerous than those which seem to have an affinity for the Pelasgian. Indeed, the language in which it is written appears almost entirely to have lost the Pelasgian element. The same observation may be made with respect to the Cortonian inscription:
Arses verses Sethlanl tephral ape termnu pisest estu;
i.e.
Avertas ignem Vulcane victimarum carne post terminum piatus esto;
or,
Avertas ignem Vulcane in cinerem redigens qui apud terminum piatus esto.
Probably, therefore, both these belong to a period at which the old Umbrian of the conquered tribes had been exercising a long-continued influence in corrupting the pure Pelasgian of the conquerors.
One example of the Etruscan alphabet is extant. It was discovered in a tomb at Bomarzo, by Mr. Dennis, inscribed round the foot of a cup, and probably had been a present for a child. The letters ran from left to right, and are as follows:
It will be seen from this specimen that the Etruscan language was deficient in the letters Β Γ Δ Ξ Ψ Η Ο Ω.
The following is a catalogue of those Etruscan words which have been handed down to us, together with their Latin interpretation. The list is but a meagre one, but valuable as containing some which have been admitted into the Latin, and as exhibiting many affinities to the Pelasgian:
Æsar Deus
Aggeltor Pater
Ainder Divus
Anthar Atrium
Astar Aquila
Apruna Haruspex
Avils Anni
Avee Vorsa Avorto ignem
Ataison Vitis
Suris Manubrium
Hutlon Sol
Capra Capra
Cassis Cassis
Celer Celer
Cipa Ego
Druma Principium
Falando Coelum
Cups Gurus
Hister Lictor
Iduce Idus
Larth Latinus
Lituus Lituus
Lortia Lanista
Laris Lar
Laucino Lucumo
Mantisa Additamentum
Naias Vagus
Noros Luxuriosus
Rasena Etrusci
Suthi Tumulus
Sulan Sol
Soe Filius
Tul Est
Tsi avil Vixit annos
Curius Ludio
Divido Divido
Ovis Idus
Vaticinamentum Vaticinor
Domitina Domina
Princeps Princeps
Additamentum Additamentum
Vagus Vagus
Luxuriosus Luxuriosus
Etrusci Etrusci
Tumulus Tumulus
Sol Sol
Filius Filius
Vixit annos Vixit annos.
The discoveries of General Galassi and Mr. Dennis at the Etruscan city of Cervetri have shown to what an extent the Pelasgian element prevailed in the old Etruscan. Cervetri was the old Cære or Agylla, which was founded by Pelasgians, maintained a religious connexion with the Greeks as a kindred race, and permitted Pelasgian to be a late period. In the royal tomb discovered in this place the name of Tarquin:
occurs no less than thirty-five times. On a little cruet-shaped vase, like an ink-bottle, was found inscribed the syllables Bi, Ba, Bu, &c., as in a horn-book, and also an alphabet in the Pelasgian character. These characters are almost identical with the Etruscan. Again, General Galassi found here a small black pot, with letters legibly scratched, and filled with red paint. Lepsius pronounced them to be Pelasgian, divided them into words, and arranged them in the following lines, which are evidently hexametrical:
Mi ni kethu ma mi mathu maram lisiai thipureuai
Ethe orai sic epana mi nethu nastav helephu.
Mr. Donaldson has offered some suggestions, with a view to explaining this inscription, and has clearly shown many close affinities to the Greek; but there is another which he quotes, and which is pronounced by Müller to be pure Pelasgian, which even in its Pelasgian form is almost Greek:
Mi kalairu fuius.
εἰμι Καλαίρου Fυιος.
It would be impossible in this work to attempt the analysis of all the known Etruscan words, and to point out their affinities to the Pelasgian, the Greek, or the Latin; but a few examples may be given, whilst the reader, who wishes to pursue the subject further, is referred to the investigations of the learned author of the “Varronianus.”
Aifil, age, is evidently from the same root as the Greek αἰών, the digamma, which is the characteristic of the Pelasgian, as it was of the derivative dialect, the Æolic, being inserted between the vowels. Aruns, an agriculturist, contains the root of ἀρόω, to plough. Capys, a falcon, that of capio, to catch. Cassis (originally capsis), that of caput, the head. Lituus, a curved staff, that of obliquus. Toga, that of tego, the dress, which was originally as much the Etruscan costume as it subsequently became characteristic of the Roman. Lastly, it is well known that, whereas the Greeks denoted numbers by the letters of the alphabet, the Romans had a system of numeral signs. This was a great improvement. The Greek system of notation was clumsy, because in reality it only pointed out the order in which each number stands. The Roman notation, on the other hand, represented arithmetical quantity, and even the addition and subtraction of quantities; and this elegant contrivance the Romans owed to the Etruscans. Their numerals were as follows :—
From the few examples which have been here given it is evident that the Pelasgian element of the Etruscan was most influential in the formation of the Latin language, as the Pelasgian art and science of that wonderful people contributed to the advancement and improvement of the Roman character.
THE OLD LATIN LANGUAGE
The above observations, and the materials out of which the old Latin was composed, have prepared the way for some illustrations of its structure and character. The monuments from which all our information is derived are few in number: the conflagration of Rome destroyed the majority; the common accidents of a long series of years completed the mischief. Almost the only records which remain are laws, ceremonials, epitaphs, and honorary inscriptions.
An example of the oldest Latin extant is contained in the sacred chant of the Fratres Arvales. The inscription which embodied this Litany was discovered A.D. 1778, whilst digging out the foundations of the sacristy of St. Peter’s at Rome. The monument belongs to the reign of Heliogabalus; but although the date is so recent, the permanence of religious formulæ renders it probable that the inscription contains the exact words sung by this priesthood in the earliest times.
The Fratres Arvales were a college of priests, founded, according to the tradition, by Romulus himself. The symbolical ensign of their office was a chaplet of ears of corn (spicea corona), and their function was to offer prayers in solemn dances and processions at the opening of spring for plenteous harvests. Their song was chanted in the temple with closed doors, accompanied by that peculiar dance which was termed the tripudium, from its containing three beats. To this rhythm the Saturnian measure of the hymn corresponds; and for this reason each verse was thrice repeated. The hymn contains sixteen letters: s is sometimes put for r, ei for i, and f for b or ph. The following is a transcription of it, as given by Orellius, to which an interpretation is subjoined:
Enos Lases juvate.
Nos Lares juvate.
Us O Lares help.
Neve luaerve Marmar sins incurrere in pleoris.
Neve luem Mars sinas incurrere plus.
Nor the pestilence O Mars permit to invade more.
Satur fufere Mars limen Salista berber.
Satiatus furendo Mars lumen Solis sta fervere.
Satiated with fury, O Mars the light of the sun stop from burning.
Semunis alternei advocapit conctos.
Semihemones alterni ad vos capite cunctos.
Us half-men in your turns to you take all.
Enos Marmor juvato.
Nos Mars juvato.
Us Mars help.
Triumpe, triumpe, triumpo, triumpe, triumpe.
Triumph, &c.
Of the Salian hymn (Carmen Saliare), another monument of ancient Latin, the following fragments, preserved by Varro, are all that remain, with the exception of a few isolated words:—
- Cozeulodoizeso, omina vero ad patula coemisse
Jam cusiones, duonus ceruses dunzianus vevet.
This has been corrected, arranged in the Saturnian metre, and translated into Latin by Donaldson as follows:—
Choroi-aulodos eso, omina enim vero
Ad patula’ ose misse Jani cariones.
Duonus Cerus esit dunque Janus vevet.
Choroio-aulodus ero, omina enim vero ad patulas aures miserunt Jani curiones. Bonus Cerus erit donec Janus vivet.
I will be a flute-player in the chorus, for the priests of Janus have sent omens to open ears. Cerus (the Creator) will be propitious so long as Janus shall live.
(2.) Divum empta cante, divum deo supplicaute.
i.e. Deorum impetu canite, deorum deo suppliciter canite.
Sing by the inspiration of the gods, sing as suppliants to the god of gods.
The Leges Regiæ are generally considered as furnishing the next examples, in point of antiquity, of the old Latin language; but there can be little doubt that, although they were assumed by the metrical traditions to belong to the period of the kings, they belong to a later historical period than the laws of the Twelve Tables. Some fragments of laws, attributed to Numa and Servius Tullius, are preserved by Festus in a restored and corrected form, and, therefore, it is to be feared that they have been modernized in accordance with the orthographical rules of a later age. One of these laws is quoted by Livy as put in force in the trial of the surviving Horatius for the murder of his sister when he returned, as the tradition relates, from his victory over the Curatii. Another is alluded to by Pliny, which forbids the sacrificing all fish which have not scales; but they are given in modern Latin, and can only be restored to their old form by conjecture.
We may, therefore, proceed at once to a consideration of the Latin of the Twelve Tables, of which fragments have been preserved by Cicero, Aulus Gellius, Festus, Gaius, Ulpian, and others. These fragments are to be found collected together in Haubold’s Institutionum Juris Romani privati lineamenta and Donaldson’s Varronianus. The laws of the Twelve Tables were engraven on tablets of brass, and publicly set up in the Comitium, and were first made public in B.C. 449. Nor had the Romans any other digested code of laws until the time of Justinian. The following are a few examples of the words and phrases contained in them:—
Ni nec Diffensus esto differatur
Em eum Se sine
Endo jacito injicito Venom-dint venum det
Aveit exit Estod esto
Fuat esto Escit est
Sonticus nocens Legassit, &c. Legaverit
Hostis Hospes
The next example of the old Latin is contained in the Tiburtine inscription, which was discovered in the sixteenth century at Tivoli, the ancient Tibur. It came into the possession of the Barberini family; but it was afterwards lost, and has never been recovered. Niebuhr considers (and his conjecture is probably correct) that this monument is a Senatus-consultum, belonging to the period of the second Samnite war. The inscription is given at length in the collection of Gruter, and also by Niebuhr and Donaldson. The Latin in which it is written may be considered almost classical, the variations from that of a later age being principally orthographical. For example:—
Tiburtes is written Teiburtes
Castoria Kastorus
Advertit advortit
Dixistis deixsistis
Publicae poplicae
Utile oitile
Inducimus indoucimus
A or ab before v af
This document is followed very closely, in point of time, by the well-known inscription on the sarcophagus of L. Cornelius Scipio Barbatus, and the epitaph on his son, which are both written in the old Saturnian metre. Scipio Barbatus was the great-grandfather of the conqueror of Hannibal, and was consul in a.u.c. 456, the first year of the third Samnite war. His sarcophagus was found A.D. 1780 in a tomb near the Appian Way, whence it was removed to the Vatican. The epitaph is as follows:—
Cornelius Lucius Scipio Barbatus Gnaivod
Patre prognatus fortis vir sapiensque
Quoius forma virtutei parisuma fuit
Consol Censor Aidilis quei fuit apud vos
Taurasia Cisauna Samnio cepit
Subigit omne Loucana opsidesque abdoucit.
“Cornelius L. Scipio Barbatus, son of Cnaeus, a brave and wise man, whose beauty was equal to his virtue. He was amongst you Consul, Censor, Aedile. He took Taurasia, Cisauna, and Samnium; he subjugated all Lucania, and led away hostages.”
His son was Consul a.u.c. 495. The following inscription is on a slab which was found near the Porta Capona. The title is painted red (rubricatus):
L. Cornelio L.F. Scipio, Aidiles, Consol, Cesor.
Hone oino ploirume cosentiunt R.
Duonoro optimo fuise viro
Luciom Scipionem. Filios Barbati
Consol Censor Aidiles hic fuet
Hic cepit Corsica Aleria que urbe
Dedet tempestatebus aide mereto.
“Romans for the most part agree, that this one man, Lucius Scipio, was the best of good men. He was the son of Barbatus, Consul, Censor, Aedile. He took Corsica and the city Aleria. He dedicated a temple to the Storms as a just return.”
It is not a little remarkable that the style of this epitaph is more archaic than that of the preceding.
The consul of the year B.C. 260 was C. Duilius, who in that year gained his celebrated naval victory over the Carthaginians; the inscription, therefore, engraved on the pedestal of the Columna Rostrata, which was erected in commemoration of that event, may be considered as a contemporary monument of the language. Some alterations were probably made in its orthography at a period subsequent to its erection, for it was rent asunder from top to bottom by lightning a.u.c. 580, and is supposed not to have been repaired until the reign of Augustus, for the restoration of a temple built by Duilius was begun by that emperor and completed by Tiberius. The principal peculiarities to be observed in this inscription are, that the ablatives singular end in d, as in the words Sicehad, obsidioned; c is put for g, as in macistratos, leciones; e for i, as in navebos, ornavet; o for u, as in Duilios, aurom; classes, navium, &c., are spelt clases, numei, and quinqueremos, triremos, quinresmos, triesmos. This monument was discovered A.D. 1565, in a very imperfect state, but its numerous lacunae were supplied by Grotefend.
About sixty years after the date of this epitaph, the Senatus-consultum, respecting the Bacchanals, was passed. This monument was discovered A.D. 1692, in the Calabrian village of Terra di Feriolo, and is now preserved in the Imperial Museum of Vienna.
There is scarcely any difference between the Latinity of this inscription and that of the classical period except in the orthography and some of the grammatical inflexions. The expressions are in accordance with the usage of good authors, and the construction is not without elegance. Nor is this to be wondered at when it is remembered that, at the period when this decree was published, Rome already possessed a written literature. Ennius was now known as a poet and an historian, and many of the comedies of Plautus had been acted on the public stage.
Having thus enumerated the existing monuments of the old Roman language and its constituent elements, it remains to compare the Latin and Greek alphabets, for the purpose of exhibiting the variations which the Latin letters have severally undergone.
The letters then may be arranged according to the following classification:—
1. Soft P C K or Q T
A. Mutes 2. Medial B G D
I. Consonants 3. Aspirates F (V). H
B. Liquids L, M, N, R
C. Sibilants S, X
II. Vowels A, E, I, O, U
Owing to the relation which subsists between P, B, and F or V, as the soft medial and aspirated pronunciation of the same letters, P and B, as well as F and V, in Latin, are the representatives or equivalents of the Greek F sound (φ and F), and V also sometimes stands in the place of β. For example (1), the Latin fama, fero, fugio, vir, &c., correspond to the Greek φήμη, φέρω, φεύγω (Fv’Ap)^. (2). Nebula, capio, albus, ambo, to νεφέλη, λαβή, αλφός, αμφω. Similarly, duonus and duellum become bonus and bellum; the transition being from du to a sound like the English w, thence to v, and lastly to b. The old Latin c was used as the representative of its corresponding medial G, as well as K; for example, magistratus, legiones, Carthaginienses were written on the Columna Rostrata, leciones, macistratus, Cartacinienses. The representative of the Greek κ was c; thus caput stands for κεφαλή: the sound qu also, as might be expected, from its answering to the Greek koppa (Q), and the Hebrew koph (ק), had undoubtedly in the old Latin the same sound as C or K, and, therefore, quatio becomes, in composition, cutio; and quojus, quoi, quolonia become, in classical Latin, cujus, cui, colonia. This pronunciation has descended to the modern French language, although it has become lost in the Italian. A passage from the Aulularia of Plautus illustrates this assertion, and Quintilian also bears testimony to the existence of the same pronunciation in the time of Cicero.
The aspirate H is in Latin the representative of the Greek χ, as, for example, hiems, hortus, and humi correspond to χείμων, χόρτος, χαμαί, whilst the third Greek aspirated mute Θ becomes a tenuis in the mouths of the early Latins, as in Cartaginienses, and the h sound was afterwards restored when Greek exercised an influence over the language as well as the literature of Rome.
The absence of the th sound in the old Latin is compensated for in a variety of ways; sometimes by an f, as fera, fores, for θήρ and θύρα.
The interchanges which take place between the T and D, and the liquids L, N, R, can be accounted for on the grammatical principle, which is so constantly exemplified in the literal changes of the Semitic languages, that letters articulated by the same organ are frequently put one for the other. Now D, T, L, N are all palatals, and in the pronunciation of R also some use is made of the palate. Hence we find a commutation of r and n in δῶρον, donum; cereus, ceneus; of t and l in θώρηξ and lorica; d and l in olfacio and odore facio, Ulysses and Ὀδυσσεύς; r and d in auris and audio, arfuisse, and adfuisse.
To the remaining liquid, m, little value seems to have been attached in Latin. In verse it was elided before a vowel; in verbs it was universally omitted from the first person of the present tense, although it was originally its characteristic, and was only retained in sum and inquam: it was also omitted in other words, as omne for omnem; and Cato the Censor was in the habit of putting dice and facie for dicam (or dicem) and faciam (faciem).
As the Roman x was nothing more than a double letter compounded of g or c and s, as rego, regsi, rexi; dico, dicsi, dixi, the only consonant now remaining for consideration is the sibilant s. The principal position which it occupies in Latin is as corresponding to the aspirate in Greek words derived from the same Pelasgic roots. Thus sus, sex, silva, &c., are represented by ὗς, ἕξ, ὕλη, &c. This may possibly be accounted for by the fact that S is in reality a very powerful aspirate. It is only necessary to try the experiment, in order to prove that a strong expiration produces a hissing sound. Those words which in classical Greek are written without an aspirate, such as εἷ, γέρων, &c., which, nevertheless, have an s in Latin, as si, senex, &c., may possibly have been at one period pronounced with the stronger breathing. The most remarkable change, however, which has taken place with respect to this letter, in the transition from the old to the classical Latin, is the substitution of r for s. Thus Fusius, Papisius, eso, arbos, &c., become Furius, Papirius, ero, arbor, &c.
The following table exhibits the principal changes undergone by the vowels and diphthongs:—
In modern Latin In ancient Latin
E was represented by i, sometimes u, as luci, condumnari
I u, ei, e, o optume, nominus, preivatus,
dedit, senatuous
U oi, ou, o quoius, ploirume, douco, honc
AE ai Aidiles
OE oi proilium
The vowels were sometimes doubled, as leegi, luuci, haace.
In the grammatical inflexions, the principal difference between the old and the new Latin is, that in nouns the old forms were longer, and assumed their modern form by a process of contraction, and that the ablative ended in d, as Gnaivod, sententiad; consequently the adverbial termination was the same as suprad, bonod, malod. The same termination appears in the form of tod in the singular number of the imperative mood.
Source: Browne, R.W. History of Classical Literature (1884)
