Martianus Capella
The Marriage of Philology and Mercury

/64/

Book III
Grammar

→ lat [221] Once again in this little book the Muse prepares her ornaments and wants to tell fabricated stories at first, remembering that utility cannot clothe the naked truth; she regards it as a weakness of the poet to make straightforward and undisguised statements,¹ and she brings a light touch to literary style and adds beauty to a page that is already heavily colored.²

→ lat [222] “But,” I cried, “in the previous book notice is given that the myths have been put away and that the precepts in the volumes which follow are a work of those Arts which tell that which is the truth.”

But with a laugh she joked at this and said: “Let us tell no lies, and yet let the Arts be clothed. Surely you will not give the band of sisters naked to the bridal couple? Surely thgy will not go like that before the senate of the Thunderer and the heavenly gods? To say no more about embellishment, what is to be the program?”

“Surely let them speak on their own teachings, and let them be clothed in incorporeal utterance.”

“Now you are deceiving me and are not consistent with your promise; why do you not admit that your work cannot be composed except by the use of imagery?” With these words the Muse got the better of me: “Are you running away?” “I am joining in the game.”

→ lat [223] So Latona’s son [Apollo] moved forward from her former place one of the servants of Mercury, an old woman indeed but of great charm, who said that she had been born in Memphis³ when /65/ Osiris was still king; when she had been a long time in hiding, she was found and brought up by the Cyllenian [Mercury] himself. This woman claimed that in Attica, where she had lived and prospered for the greater part of her life, she moved about in Greek dress; but because of the Latin gods and the Capitol⁴ and the race of Mars and descendants of Venus, according to the custom of Romulus [i.e., Roman custom] she entered the senate of the gods dressed in a Roman cloak.⁵ She carried in her hands a polished box, a fine piece of cabinetmaking, which shone on the outside with light ivory, from which like a skilled physician the woman took out the emblems of wounds that need to be healed. → lat [224] Out of this box she took first a pruning knife with a shining point, with which she said she could prune the faults of pronunciation in children; then they could be restored to health with a certain black powder⁶ carried through reeds, a powder which was thought to be made of ash or the ink of cuttlefish. Then she took out a very sharp medicine which she had made of fennelflower and the clippings from a goat’s back, a medicine of purest red color, which she said should /66/ be applied to the throat when it was suffering from bucolic ignorance and was blowing out the vile breaths of a corrupt pronunciation.⁷ She showed too a delicious savory, the work of many late nights and vigils, with which she said the harshness of the most unpleasant voice could be made melodious. → lat [225] She also cleaned the windpipes and the lungs by the application of a medicine in which were observed wax smeared on beechwood and a mixture of gallnuts and gum and rolls of the Nilotic plant (papyrus]. Although this poultice was effective in assisting memory and attention, yet by its nature it kept people awake. → lat [226] She also brought out a file fashioned with great skill, which was divided into eight golden parts joined in different ways, and which darted back and forth—with which by gentle rubbing she gradually cleaned dirty teeth and ailments of the tongue and the filth which had been picked up in the town of Soloe.⁸

→ lat [227] She is reckoned to know by the effort of frequent calculations these arcane poems and manifold rhythms. Whenever she accepted pupils, it was her custom to start them with the noun. She mentioned also how many cases could cause faults or could be declined accurately. Then, appealing to her pupils’ powers of reasoning, she firmly held the different classes of things and the words for them, so that the pupils would not change one name for another, as often happens with those who need her attention. Then she used to ask them the moods of the verbs and their tenses and the figures,⁹ and she ordered others, on whom complete dullness and inert laziness had settled, to run through the steps and to climb upon as many works as possible, treading on the prepositions or conjunctions or participles, and to be exercised to exhaustion with every kind of skill. → lat [228] Some of the gods called this woman, so clever in so many ways, Iatrice,¹⁰ others called her Genethliace,¹¹ because /67/ of the different aspects of her work; she is also skilled at restoring the truth, and neither Pallas nor Maia’s son himself would deny that she can assist in faults of pronunciation. Yet it seemed inappropriate that she should enter as a doctor in Roman dress. So she was asked for her name and her profession and an explanation of her whole field of study.

→ lat [229] As if it was normal for her to explain what had been asked and easy to tell what was wanted of her, she modestly and decently folded back her cloak from her right hand and began: “In Greece I am called Grammatice, because a line is called grammé and letters are called grammata and it is my province to form the letters in their proper shapes and lines. For this reason Romulus¹² gave me the name Litteratura, although when I was a child he had wanted to call me Litteratio, just as amongst the Greeks I was at first called Grammatistice; and then Romulus gave me a priest and collected some boys to be my attendants. Nowadays my advocate is called Litteratus, who was formerly called Litterator. This is recalled by a certain Catullus, a poet not without charm, when he says ‘The Litterator Sylla renders you his service.’¹³ Such a man was called by the Greeks Grammatodidaskalos.

→ lat [230] My duty in the early stages was to read and write correctly; but now there is the added duty of understanding and criticizing knowledgeably.¹⁴ These two aspects seem to me to be shared with the philosophers and the critics. Two of these four functions may be called active, and two contemplative, since indeed we are active when we write or read anything, but we are engaged in the contemplation of the result when we understand or assess what has been written, although these four functions are all linked by a certain /68/ affinity, just as is shown to happen in other studies also. For the actor understands at the beginning what is the efficacy of acting, and the astronomer does certain things in order to know through them what he ought to demonstrate. The geometrician combines those functions, in that he both writes and comprehends the shapes of the theorem, assisted by sure calculation.

→ lat [231] I have four parts: letters, literature, the man of letters, and literary style. Letters are what I teach, literature is I who teach, the man of letters is the person whom I have taught, and literary style is the skill of a person whom I form. I claim to speak also about the nature and practice of poetry. Nature is that from which speech is formed. Practice occurs when we put that material into use. To these we add the matter, so as to know what we must talk about. Speech itself is taught in three steps; that is, from letters, syllables, and words.¹⁵

→ lat [232] In respect of letters, two questions arise. Letters are either natural or artificial. The names of letters are formed into speech by the operation of nature; artificial formation has laid down their written forms to the end that people in each other’s presence could use one form, those absent another.¹⁶ From the point of view of the writer they are called mute, from the reader’s they are called voiced, /69/ if indeed the latter can be taken in only by hearing and the former only by sight.¹⁷

→ lat [233] There are some letters which can form a complete sound by themselves, others which by themselves can form none. For there are the vowels, of which the Greeks say there are seven, early Latin¹⁸ says six, later usage five – rejecting y, as a Greek letter. These vowels in Latin can be long or short, acute, grave, or even circumflex in accent, combined and separated without losing their names. Sometimes single vowels form syllables, sometimes thev accept consonants and certain vowels on either side, sometimes thev change themselves, in recent times they follow themselves with pleasing effect. For instance, a often changes into e, as capio gives cepi; sometimes into i, as in salio, insilio; sometimes into o, as in plaustrum, plostro; or into u,¹⁹ as in arca, arcula. Le letter e changes first into a, as in sero, satum; or into i, as in moneo, monitus; or into o, as from tegendo comes toga; or into u, as from tego you get tugurium. Similarly, the vowel i changes into a, as in siquis, siqua; into e, as in fortis, forte; into o, as in qui, quo; into u, as in ibi, ubi. In the same way the letter o changes into a as in creo, creavi; or into e, as in tutor, tutela; or into i, as in virgo, Virginis; or into u, as in volo, volui. Again, u likewise changes into a, as in magnus, magna; into i, as in telum, teli; into o, as in lepus, leporis; into e, as in sidus, sideris. The changed letters can be linked with letters on either side, or on one side only, or on none; and sometimes they end some parts of speech, and sometimes they do not.

→ lat [234] To begin with, a takes the letters v and i on either side of itself: we say aurum and varus and Ianus and Aiax. It takes e on one side only, as Aeneas; but it does not take o beside it at all. It ends masculine nouns, like Iugurtha; feminine, like dea; and neuter nouns in the singular (but only Greek ones) like toreurna, peripetasma; and in the plural, Latin nouns like monilia; in verbs it ends the imperative mood, as in canta, salta. → lat [235] The vowel e has the force of two Greek letters; when it is short, it is the Greek epsilon, as /70/ in ab hoc boste; and when it is long, it is the Greek eta, as in ab hac die. Then especially it takes a circumflex accent. It takes the two vowels i and v on either side, as in Euro, Veientano, and eia, iecore. It ends neuter nouns, such as monile, and Greek feminine nouns such as Calliope, and pronouns such as ille, iste; it ends numerals of every gender, such as quinque, although the numeral is indeclinable; it also ends verbs in the imperative mood, such as sede, curre, and in the infinitive mood, such as scribere, scripsisse. → lat [236] The letter i takes all vowels on cither side: it takes a, as when we say ianuariae and Aiax; the examples given above have shown how it takes e; it is joined to o in the names Iovis and Oinone, and oisus is also used—that is the archaic form of usus; and it takes v, as in iuvando and vita. Sometimes this vowel makes a double syllable;²⁰ it ends nouns of every gender, such as frugi; and numerals of every gender, such as viginti; and verbs in the indicative mood, such as novi, memini; in the imperative, such as sali, veni; and in the infinitive, such as iaculari, luctari. → lat [237] I have noted above with what vowels the letter o is juxtaposed, and in what positions. It ends masculine nouns, such as Cato; feminine nouns, such as Iuno; numerals of all genders, such as octo; the active verb, such as canto, laudo (it is found in the imperative mood only in the form cedo); and it can end an adverb, such as subito; or a preposition, such as pro. → lat [238] The letter v is juxtaposed with every other vowel on either side, except that o does not precede it. V [u] alone of vowels, when it occurs twice consecutively, does not form a double vowel but conjoins as if with a consonant, as in vulgus, vulcanus. It ends articles,²¹ such as tu, and neuter nouns like genu, cornu; but it docs not end any verb.

→ lat [239] So much for vowels. I was instructed to expound all the letters, not only the vowels. It is beyond dispute that the remaining letters, which now follow, are consonants. They are divided into semivowels and mutes; I shall set out the list of each. → lat [240] The /71/ letter f, which is the first of the semivowels, precedes only two consonants, l and r, as in flavus and frugi. In juxtaposition with vowels, it will not take i and v preceding it, and it is never the final letter of a word.²² → lat [241] The letter l has three different sounds. It gives a thin sound when it is doubled, as in sollers, Sallustius; a sound of moderate degree when it ends nouns, as in sol, sal; again it gives a light sound when it precedes vowels, as in lapis, lepus, liber, locus, lucerna; and a full sound²³ when it is preceded by the letters p, g, c, f, as in Plauto, glebis, Claudio, Flavo. L never precedes any semivowel or mute. It substitutes for d in a prepositional prefix when a precedes it, as in allidit, aliigat.²⁴ When the preposition in is prefixed, the l is kept, as in illepidus, illotus.²⁵ It takes vowels on either side, and it is the final letter of some nouns: in the masculine sol; in the feminine Tanaquil; in the neuter when a, e, i precede it, as in bidental, mel, sil. It is also the final letter of adverbs, as in semel.

→ lat [242] The letter m can equally well be at the beginning, the middle, or the end of words, as in mores, umbra, triticum. It precedes only n amongst the consonants, as in Memnon; and it is preceded only by s, as in Sminthius; and it is sometimes changed into n, as when we say nunquis.²⁶ It is the final letter of neuter nouns like aurum and argentum, and of indeclinables like nequam, and of the masculine gender’s accusative singular and genitive plural,²⁷ and of participles, verbs, and adverbs, such as tractum, legam, cursim; and of numerals of all genders, such as novem, decem.

→ lat [243] The letter n receives a fuller value as the first or the last of a word, as in Nestor, tibicen; and a slighter value in the middle of a word, as in mane, damnum. Amongst the consonants, it follows the letter m, as in Memnon; and it precedes s, as in fons, mons. It is assimilated into m when , p, or m follow it, so that we say imbuit, /72/ impulit, imminet.²⁸ It is the final letter of some nouns: masculine, such as tibicen; Greek feminine nouns, such as Siren; neuters, such as culmen.

→ lat [244] The letter r accepts vowels on either side of it. Of the consonants, it precedes only s, as in sors, fors; and occasionally the double letter x, as in arx. It is converted into l and n and s, as niger into nigellus, femur into feminis, gero into gessi. It can be the final letter of masculine nouns after any vowels, as in par, pater, vir, nitor, fur; of feminine nouns after e and o, as in mulier and soror; and of neuter nouns after all vowels except i, as in calcar, piper, marmor, sulphur. It can be the final letter of verbs and adverbs, as in venor, vador, ter.

→ lat [245] Some people have not regarded s as a letter; Messala said it was a kind of hiss. It accepts the letter p on either side, as in spado, psittacus; and it accepts c, q, and t after it, as in Scaurus, squama, Stella. It is more often converted into other letters, such as l when we say modus and modulus; or n in the case of sanguis, sanguinis; or r in flos, floris; or d in custos, custodis; or t in nepos, nepotis. It can be the final letter of masculine and feminine nouns after any vowel and of neuter nouns after a, o, u, as in vas, os, nemus. To s the emperor Claudius added p or c, because of the Greek letters psi and xi, as in psalterium, sacsa (saxa). → lat [246] No one thinks x is a letter, because it is a double letter. It is made up of g and s, as in rex, regis; or of c and s as in pix, picis, or nux, nucis. It sometimes changes into v, as in nix, nivis; or into ct, as in nox, noctis. It is never an initial letter in Latin, but may be in Greek, as in Xanthus. It is juxtaposed with anv vowel in masculine nouns, such as Aiax, frutex, calix, velox, Pollux, and in feminine nouns, such as fax, lex, lodix, celox, lux; it is never the final letter of a neuter noun.

→ lat [247] Having dealt with the semivowels, we must examine the mutes. These are so called because, unless vowels occur in association with them, their sound ceases at its beginning within the mouth. → lat [248] B, first of all, accepts any vowel on either side of it, and it precedes the consonants l and r, as in blaesus and brevis. It is converted into c, as in succurrit; into f, as in sufficit; into m, as in /73/ summittit; into p, as in supponit; into s, as in sustulit.²⁹ It is never a final letter, except of the three prepositions, ab, ob, sub. → lat [249] C accepts vowels on either side of it, and it precedes certain consonants; namely, l, t, r, m, n, as in clarnus, tectum, crus, Acmon, Cnidus. It is the final letter of the pronouns³⁰ hic, haec, hoc, and of adverbs, such as sic, huc, hic. C alone of the mutes lengthens the preceding vowel and is treated as a double letter, as in hic, hoc, hac.³¹ → lat [250] D accepts vowels on either side of itself. It precedes the letter r in Drusus, but never precedes the letter m in Latin, though it sometimes does in Greek, as in Dmolus; also n, as in Ariadne. It is converted into c, as in accidit; into g, as in aggerat; into l, as in allegat; into p, as in apponit; into r, as in arripit, into s, as in assidet; into t, as in attinet.³² It may be the final letter of neuter pronouns, such as istud, illud; and of prepositions, as in apud. → lat [251] G precedes any vowel and follows the letter a, as in aggere; whenever g is doubled, a is preceding it. G precedes the letter r, as in grave; l, as in gladius; n, as in ignis. It changes into c, as in rego, rector. It is never a final letter. → lat [252] It is beyond doubt that h is the symbol of the aspirate; when it occurs it is an addition to a vowel, as in hospes and heres. It changes into x, as in traho, traxi. The Greeks divided it into two – its right half the sign of the aspirate, its left signifying the contrary.³³ → lat [253] K is thought sometimes to be a notation, sometimes /74/ a letter, for the letter c can certainly perform all its functions except these: kapita, kalendae, kalumniae.

→ lat [254] P accepts vowels on either side of itself, and precedes the letter r, as in prandere; l, as in placere; t, as in Ptolemaeus; s, as in psittacus; n, as in Sipnus; it is never a final letter. → lat [255] Some people used not to consider q to be a letter, and they would be right, if only it were not so evident in aequus and equitatus. It never occurs without two vowels in association, of which the first is always u; the other vowels may then follow, as in quartus, questus, Quirites, quotus, equus. It is made up of c and u and so is a double and composite sound, and double-q does not occur. It is not preceded by any particular letter,³⁴ and it is never a final letter. → lat [256] T accepts vowels on either side of itself, as in coniunctiones and praepositiones. It precedes l in Tlepolemus, m in Tmolus, n in potnia, r in Troia. It can be the final letter of neuter nouns, such as caput, sinciput, lact. It ends articles,³⁵ such as quot, tot; verbs, such as legunt; an adverb and an interjection, such as ut, attat. → lat [257] Z comes from the Greeks, although they themselves at first used the Greek sigma; they used to say Sethum, which now they call Zethum; they also consider it a double letter. It is made up of t and s, and being obviously a double letter it never occurs twice consecutively. It precedes the letter m, as in Zmyrna. → lat [258] Out of all these, 18 letters meet all the requirements of a piece of composition. I am satisfied to include y amongst the vowels; without it one cannot write Hyacinthus or Cyllenius. This gives us a total of six vowels and 12 semivowels and mutes. → lat [259] For if h is treated as an aspirate symbol; and q and k are unnecessary; and x, being a double letter, lacks the simplicity prerequisite for the basic letters; and z is excluded from Latin; then, as I said, 18 letters remain.

→ lat [260] All these letters, including those classed as unnecessary or not genuine, are formed by the sound of any single voice; and in /75/ the harmony of the voice they have uncovered the various elements of the concord of nature.

• → lat [261] We utter A with the mouth open, with a single suitable breath. We make B by the outburst of breath from closed lips.
• C is made by the back teeth brought forward over the back of the tongue.
• D is made by bringing the tongue against the top teeth. E is made by a breath with the tongue a little depressed. F is made by the teeth pressing on the lower lip.
• G, by a breath against the palate.
• H is made by an exhalation with the throat a little closed.
• I is made by a breath with the teeth kept close together.
• K is made with the palate against the top of the throat. L is a soft sound made with the tongue and the palate.
• M is a pressing together of the lips.
• N is formed by the contact of the tongue on the teeth.
• O is made by a breath with the mouth rounded.
• P is a forceful exhalation from the lips.
• Q is a contraction of the palate with the mouth half-closed.
• R is a rough exhalation with the tongue curled against the roof of the mouth.
• S is a hissing sound with the teeth in contact.
• T is a blow of the tongue against the teeth.
• U is made with the mouth almost closed and the lips forward a little.
• X is the sibilant combination of C and S.
• Y is a breath with the lips close together.
• Z was abhorrent to Appius Claudius, because it resembles in its expression the teeth of a corpse.

→ lat [262] The rules of literacy, the primary study, which usually are spread over many volumes, are quickly brought into a small compass: in it we see what any letter connects together by its association on either side; on which side it accepts or demands an accompanying letter; how a letter customarily is altered by the laws of transformation and thus accepts a different name; what sounds of the mouth or movements of the tongue or outbursts from the lips form the letters. Now we must survey the syllable, composed of letters in combination: how it is accented; when it is long or short. I shall survey these two points, since the preceding study of the letter has at the same time taught the connections. /76/ The order of studies takes this as the next matter to be dealt with, if that meets with your divine approval.”

→ lat [263] While Grammar was saying this, and Jupiter and the Delian were urging her forward, Pallas spoke up: “While Literature here is hurrying on to discuss the connection of syllables, she has passed over the historical aspect.” At this objection by the maiden goddess, Grammar in great agitation answered: “I know I must pass over a great deal, so as not to incur the distaste of the blessed by getting entangled in details. So I shall perform my purpose, hastening along the shortest ways, to avoid getting lost, hidden in thick undergrowth or a dense mass of briars.

→ lat [264] A syllable is a combination of letters which makes and is heard as a single sound. It is agreed, as I said, that it has three aspects: its connection, its accentuation, and its length. → lat [265] There are four kinds of connection, two natural and two historical. They are natural when each syllable is examined to see whether it is joined on one side or on both or on neither, and whether letters which cannot by themselves come together may be united when some other letter is combined with them – as with m and n, which cannot combine of themselves, but can when a vowel is added, as irt amni and somno – we may pass over these kinds of connection, since they have been dealt with in our discussion of letters → lat [266] The historical type of connection occurs in dealing with letters which can be combined of themselves, when different modes of speech join them differently and with different numbers of letters. For we write the first syllable of Musae with two letters, but the Greeks write it with three.³⁶ Lucilius joins a and e in the dative case, saying huic Terentiae, Orbiliae; but in the genitive he joins a and i, and so does Lucretius often; and Vergil, as in aulai, pictai.³⁷ → lat [267] There are, moreover, two kinds of connection when we join syllables in a manner to fit a rhythmical foot or verse, so that two letters suddenly vanish into one, as in synaliphe (for synaloephe), or when certain letters are dropped from a word, as in ec(th)lipsis.

→ lat [268] So much for connections; now let us look at accentuation, a topic which the Greeks call peri prosodion [pitch]. This has three divisions: each and every syllable is either grave or acute or /77/ circumflex;³⁸ and just as no sound exists without a vowel, so is none without an accent. The accent, as some have thought, is the soul of a sound, the seedbed of music, because every melody comes from the accents and weight of sounds; and so it has been called ‘accent’ from ad cantus [tuning]. → lat [269] Every utterance in Latin, whether simple or compound, has one sound that is either acute or circumflex; none can ever have two acute or circumflex sounds, but two grave sounds are common. The acute is on the first syllable when you say Caelius, on the second in Sallustius, on the third in Curiatius, and a circumflex on the first when you say caelum; the penultimate is never acute when short by nature, but the preceding syllable, the third from the end, may be acute whether short or long, as in Cicero, Caelius. The circumflex stands only on the penultimate syllable, and only when that syllable is long by nature and the final syllable short, as in Galenus. If the final syllable is long, or the penultimate is long by position and not by nature, even when the final syllable is short, the penultimate bears an acute accent, as in Galeni, Camilli. It is a peculiarity of the circumflex accent that it falls only on syllables long by nature, whereas the acute accent is found on both long and short syllables. Every voiced sound must have either an acute or a circumflex accent, even if the sound be a monosyllable; monosyllables do not have the grave accent. Every monosyllabic word, if it is short or long by position, has the acute accent, as in far, ars; if it is long by nature, it bears the circumflex, as in lux, mos.

Disyllabic words have an acute accent on the first syllable when both syllables are short, as in citus; or when both are long by position, as in sollers; or when the final syllable is long by position, as in cohors. If, however, the first syllable is long by nature and the final syllable short, then the first syllable bears the circumflex, as in lima; if the final syllable is long by position or by nature, the first syllable bears the acute accent, as in codex, docte; in a disyllabic word the grave accent will never be found on the first syllable.

In a trisyllabic word, if the middle syllable (which we call the /78/ penultimate) is short, it will always be pronounced with a grave accent, and the syllabic immediately before it, the third from the end, will have an acute accent, as in Catulus; but if the middle syllable is long, its accent depends on whether it is long by nature or by position. If it is long by nature and the last syllable is short, the middle syllable has the circumflex, as in Cethegus, Mancinus; if the middle syllable is long by nature and the final syllable is long, the middle syllable has the acute accent, as in Catoni, Ciceroni; if the middle syllable is long by position it will have the acute accent, whatever the length of the final syllable, as in Catullus, Catullo, Metellus, Metello. → lat [270] Those pronouns which may be declined in two ways --- that is, with the middle syllable of the genitive case either shortened or lengthened, like ipsius and illius --- vary their accentuation accordingly; if their second syllables are shortened, their first syllables have the acute accent, as in ipsius, illius; if their middle syllables are long, the first syllables have the grave accent, the second syllables the circumflex, as in ipsius, illius. So the first syllables have the acute accent when the middle ones are short; the middle ones when long have either acute or circumflex: acute when the last syllables are long, circumflex when these are short. → lat [271] The words tenebrae, latebrae, manipli differ from the former examples in this: that the former examples have their consecutive consonants in separate syllables,³⁹ whereas these three words, though their penultimate syllable is long by position, have no consonant at the end of the second syllable, but have two at the beginning of the third.⁴⁰ So it comes about that the middle syllable produces the nouns tenebrae and latebrae with an acute accent on the first syllable;⁴¹ but maniplos /79/ and fenestras have an acute accent on the penultimate syllable: some scholars have thought that that kind of penultimate syllable is always short, because it is not closed with a consonant; some, that it is long, because, although it is not closed with a consonant, the next syllable begins with two consonants, and because the natural softness of the letters l and r make syllables now long, now short. But no Latin word of two or more syllables has an acute accent on the final syllable; and a word never has a circumflex on the final syllable unless the antepenultimate syllable is long by nature, as in ergo and pone.⁴²

→ lat [272] The true pronunciation of individual sounds is frequently altered or even lost by their context in speech. Those sounds are lost which are carried onto a succeeding sound, as, in the case of prepositions, in the phrase ante urbem; here ante loses its acute accent on the first syllable; similarly with post muros. Words change their accents when -que, -ve, and -ne are suffixed; when -que is a copulative conjunction, and -ve an expletive one, as in Latiumque augescere vultis and stimulove meum cor in the Pelopidae of Accius. The acute accent will never move from the first syllables to the last except in the case of suffixed particles, which have the effect of rendering acute the final syllables of words to which they are suffixed.

When Greek nouns are brought into Latin, they are pronounced according to our rules unless they remain in their Greek forms. Thus, if Olympus and Caucasus are pronounced as Latin words, it is clear, from the principles set out above, on which syllables they should bear the acute or circumflex accent. But if they remain in their Greek form (in the case of foreign words, we agree to this rule), then they nevertheless must keep the acute accent on the middle syllable in the oblique cases, since we hear Greeks themselves pronouncing them in no other way. → lat [273] The acute accent is signified by a stroke ascending from the left to the right thus ‘; the grave descends from left to right, thus ‘; the symbol of the circumflex is a semicircle placed above the letters themselves ^. We sometimes /80/ call accents fastigia [gables], because they are placed on top of letters; sometimes cacumina [summits], toni [tones], or soni [sounds]; and the Greeks call them prosodiae [pitches]. It is worth knowing that all three accents can occur in a single word, such as Árgìlêtum.

→ lat [274] I think I have given a sufficient outline of what the accents are on the syllables and in what circumstances they are modified; now I shall explain the lengths of syllables. Each syllable is either short or long or common. It is short when the vowel is short and nothing following helps to make it long, or when the syllable ends with a short letter or form.⁴³

→ lat [275] A syllable is rendered long in two ways, by nature and by position: by nature when the vowel itself is long, or when⁴⁴ the syllable has a circumflex accent or an acute on the penultimate, or when the vowel is a monosyllabic word, or when the syllable contains a diphthong, or when the first syllable, of any word, even in some other compound form, does not change its vowel or accent;⁴⁵ by position, when a short vowel is followed by two consonants, cither in the same syllable or in the next syllable or shared between the two syllables, or by a double consonant in either syllable. → lat [276] The double consonants are x and z; sometimes i and c can be considered double consonants, as when i comes between vowels, as in aio, Troia; or when c does, as in hoc erat alma parens.⁴⁶ These are the types of consonant or double consonant which can make short vowels into long syllables if they occur at the end of the same word. If they occur in the following phrase, the consonants do not lengthen the preceding syllable, although Vergil would contest this, saying arma virumque cano / Troiae,⁴⁷ and again fontesque / /81/ fluviosque voco.⁴⁸ → lat [277] The diphthongs are ae, oe, au, eu, ei; there are no diphthongs without these syllables.

→ lat [278] There are eight ways in which syllables may be common. The first is when a short vowel is followed by two consonants, the second of which is a liquid: for then it is long, as in vasto Cy̆clopis in antro;⁴⁹ or short, as in vastosque ab rupe Cy̆clopas.⁵⁰ The second way is when a short vowel is followed by two consonants, the second of which is b: for then it is long, as in terga fatigamūs hasta;⁵¹ or short, as in quisquĭs honos tumuli⁵² The third way is when a short vowel is followed by two consonants, the first of which is s: for then it is long, as in undē spissa coma;⁵³ or short, as in ponitĕ spes sibi quisque,⁵⁴ The fourth way is when a short syllable ends a period: for at the break it is long, as in nam tibi, Thymbre, capūt Euandrius;⁵⁵ or short, as in hoc capŭt o cives.⁵⁶ On another occasion a diphthong can make a common syllable: long, as in musae Aonides⁵⁷ short, as in insulaĕ Ionio in magno.⁵⁸ For when it is followed by a vowel, a diphthong can become short, as in sudibusve praĕustis.⁵⁹ Again, when a long vowel is followed by another vowel, it can be long, as in Ō ego infelix, quem fugis;⁶⁰ or short, as in sub Iliŏ alto.⁶¹ Again when a pronoun ends with the letter c and is followed by a vowel, it can be /82/ long, as in hōc erat alma parens;⁶² or short, as in solus hĭc inflexit sensus.⁶³ Again, when a short vowel is followed by z, it can be long, as in Mēzenti ducis exuvias;⁶⁴ or short, as in nemorosa Zacynthos.⁶⁵

I have discussed the nature of the three kinds of syllables;⁶⁶ it remains to discuss final syllables, which are authoritative for the grammarian and define the role of the word.⁶⁷ First I must deal with the noun. → lat [279] When the nominative singular ends with a, the letter is short, as in Catilina, Iulia. When it ends with e, this is long in Greek words, as Euterpe.⁶⁸ The final letter i is long, as in frugi. The final o is short in Latin nouns, as in Cato.⁶⁹ The final u is short, as in cornu. The final al is short, as in tribunal; el is short, as in mel; il is short, as in vigil, except in the one Etruscan word Tanaquil. Ol is long, as in sol; ul is short, as in consul. When the nominative singular ends with m, the syllable is short, as in tectum, although an example of this seldom occurs, because an m between two vowels is elided to avoid mytacism.⁷⁰ The final an is long, as in Titan; en is short in neuter nominatives such as carmen, but long in other genders, as in lien, Siren. The ending on is long, as in Memnon. The ending ar is long only in monosyllables such as Nar, far. The ending er is short in Latin nouns such as puer, with the single exception of the monosyllabic ver; but in Greek nouns such as aer it is long. The ending ir is short, as in vir; the ending or is short, as in auctor. The ending ur is short, as in murmur, with the single exception of the monosyllable fur. The ending as is long in Latin nouns such as facultas, paupertas; and in Greek nouns it is short only in those instances when the genitive ends in dos, as in Arcas, Arcados. The nominative singular ending es is long in Greek nouns such as Anchises; in Latin nouns of the fifth declension it is long, as /83/ in dies; in the third declension it is long in those instances where the genitive singular has no more syllables than the nominative, such as labes, labis; or where the genitive singular has more syllables but keeps the long e in the antepenultimate syllable, as in quies, quietis. If the e of the nominative changes to i in the genitive, as in miles, militis, or if the e in the genitive is short, as in seges, segetis, then the e of the nominative is short, except in Ceres, pes, and words like these. The nominative ending is is short, as in agilis. The ending os is short in those Latin nouns in which the genitive’s antepenultimate syllable is not naturally long, such as os, ossis. In Greek words the nominative termination os is short when the genitive ends in a diphthong, as in Delos, Delou. The ending us is short, as in doctus, but it becomes long if the genitive has more syllables than the nominative and has a long u in the antepenultimate syllable, such as virtus, virtutis, and in one indeclinable word, pus. Nominatives ending in t have short final syllables, as in caput. Only two nouns end their nominatives with c, namely allec, where the final syllable is long, and lac, whose inflection is disputed. → lat [280] The genitive singular is long when it is similar to the nominative, as in senatus; or when it has a diphthong, as in Iuliae; or when it ends in i, as in docti; in other instances it is short. The dative singular is long...,⁷¹ as in Pompeio; in Greek words it is short if it ends in the letter i, as in Palladi. The accusative singular in Greek words is short only when it ends in a or on, as in Thesea, Delon; but in Latin words it is short, as in doctum. The vocative singular ending in a in Latin nouns or in feminine Greek nouns is short, as in tabula, Musa; but in masculine Greek nouns it is long, as in Aenea. The vocative ending e in Latin words is short, as in docte; in Greek nouns it is long, as in Tydide, except in those nouns which have the Greek nominative ending os, such as Phoebos, Phoebe. The vocative ending i in Latin nouns is long, as in Mercuri; in Greek nouns it is short, as in Nai. The vocative ending o in Latin nouns is short, as in Cato, although Vergil feels otherwise; in Greek nouns it is long, as in Dido. The vocative ending u is short, as in cornu; vocatives which end in consonants follow the rule for pronunciation of the nominative, with the exception of Greek nouns, in which the vocative often differs in length from the nominative because of the difference between /84/ the two languages, as in the noun Diomedes.⁷² The ablative singular is always long, except when it ends in e, at least in those nouns which are of the third declension, as in a pariete. → lat [281] The nominative and vocative plural in the masculine and feminine genders are long, as in fluctus, terrae; in neuters they are short, as in fata; in Greek nouns they are short when they end in es or a, as in rhetores, poemata; in other instances they are long, as in musae. The genitive plural ending is short, as in doctorum, though in Greek nouns following the Greek declension, such as Philaenon, it is long. The dative and ablative plural ending in is is long, as in doctis; the ending in bus is short, as in nominibus; if the noun is declined as in Greek and the dative plural ends in in, as in Arcasin, the ending is short; otherwise it is long. The accusative plural in masculine and feminine nouns is long, as in doctos, Iulias; in neuter nouns it is short, as in moenia; but in Greek nouns, if the accusative plural ending is as and the noun’s genitive singular ending is os, then the accusative plural ending is short, as in Arcados, Arcadas; otherwise it is long, as in Musas.

→ lat [282] In all pronouns the nominative singular is short, except the two monosyllables tu and qui. In the genitive singular, i or ae at the end of the word are pronounced long, as in mei, meae; otherwise they are short, as in illius. The dative singular ending is always long, as in nostro, except in mihi, tibi, sibi, which are regarded as either long or short. The accusative singular ending is short, as in illum, with the exception of me, te, se. The vocative singular is the same as the nominative singular. The ablative singular is long, as in ab illo. The nominative and accusative plural are short only when they end in a, as in nostra. The genitive plural is always short, as in illorum. The dative and ablative plural ending is is long, the ending bus is short.

→ lat [283] In all verbs, their moods, tenses, numbers, persons, and conjugations, this is the standard rule: whatever part ends in a, that ending is long, as in canta. The ending e is short, as in lege, unless it is in a verb of the second conjugation, such as doce; then it is long. A part ending in i is always long, as in nutri. The ending o is short, as in audio, although the authorities disagree on this: Vergil makes the first person singular of the first conjugation long, as in /85/ canto quae solitus⁷³ and terra tibi mando.⁷⁴ However, monosyllables such as ++doùù, ++floùù, ++stoùù are long. The ending u is long...,⁷⁵ The ending am is short, as in legebam. The ending in r is short, as in loquor. The endings in the letter s are long if the penultimate letter is a or e, as in amas, doces, with the exception of the monosyllable es and its compounds, such as ades; but if i is the letter before the s, it is short, as in legis, legitis; except that the second person singular of the present indicative in the third conjugation⁷⁶ is long, as in nutris, audis, and the same part of the verb volo, vis; if the letter before s is u, it will be short, as in nutrimus. The ending in t is short, as in legit. The ending in c is long, as in produc. Participles, although they are inflected through cases, are pronounced by principles completely different from those for the pronunciation of nouns.

→ lat [284] Monosyllabic adverbs like huc, or their compounds like illuc, are all long, except bis and ter; for other adverbs the ending in a is long, as in una; the ending in e is long, as in docte, pulchre, except in those cases where the adverb has no comparative or superlative, such as rite, or those in which the comparative is defective, such as bene,... impune.⁷⁷ The ending in i is long, as in heri, except for ibi and ubi and their compounds like sicubi, and for quasi also. The endings in o of adverbs which are not derived from other parts of speech are short;⁷⁸ if they are derived from other parts of speech, as in falso, the endings are long, although the authorities differ. Endings in u are long, as in noctu; those in l should be short, as in semel; those in m, n, r are short, as in cursim, forsan, fortiter; those in s are long only when the s is preceded by a, as in alias; those in c are long, as in illuc.

→ lat [285] In participles, the nominative singular ending is always /86/ short,⁷⁹ as in lecturus; the genitive and dative singular endings are long, as in lecturi, lecturo, except in the present tense, where the genitive ending is short, as in amantis. The accusative and vocative endings are definitely short, as in lecturum, lecture; the ablative singular ending is short only when it ends in e, as in amante, legente. The nominative and accusative plural endings in participles of the masculine and feminine genders are long, as in lecturi, lecturos, lecturae, lecturas; in neuters they are short, as in lectura. The genitive plural ending is short, as in lecturorum. The dative and ablative plural endings in is are long; those in bus are short.

→ lat [286] The endings of copulative, disjunctive, and expletive conjunctions are short unless their position makes them long. Causal and rational conjunctions ending in a, such as propterea, interea, have their endings long, except for it a and quia; those which end in i, such as si, have their endings long, with the sole exception of nisi, those which end in n have the ending long if i precedes the n, as in sin; otherwise the endings are short. AU other conjunctions are regarded as having short endings.

→ lat [287] Prepositions ending in a, such as contra, extra, and the monosyllabic a, are the only prepositions with long endings. Prepositions governing the ablative case all have short endings, except for those monosyllables which cither consist of a vowel or end in a vowel, such as e, de. Prepositions which govern more than one case are certain to have short endings.

→ lat [288] Interjections are long if they are monosyllables like heus; but if they are of two or three syllables, since they have the appearance of one of the other parts of speech, their length is determined by the standard applying to the part of speech they resemble, as in the case of papae.

These, then, are the rules to be observed in pronouncing the final syllables in all the parts of speech, except with respect to diphthongs and to words long by position.

→ lat [290] Analogy, which in Latin is called proportio, is the observance of agreement between similar words. In the first place, all Latin nouns end in one of twelve letters: the five vowels, the six semivowels, and one mute, t, as in caput.

A, the first of the vowels, ends masculine nouns such as Catilina, Iugurtha, feminine nouns...⁸⁰ such as advena, Greek feminines such as Helena, Andromacha, and pronouns such as altera, sola, illa. In these, if you substitute ius for the a ending, you get the genitive singular, alterius, solius, illius; if you substitute i, you get the dative, alteri, soli, illi. Greek neuters like poeina, toreuma, have identical endings in three cases of each number; they add tis to form the genitive singular, omit the s to form the dative, and end the ablative with a short e; in the dative and ablative plural they keep the form of the genitive singular, although they can also add the bus ending.

→ lat [291] Those Latin nouns which end with a short e, such as monile, sedile, are neuter. Besides those cases which have identical endings (as being neuter)⁸¹ these words have identical dative and ablative endings (huic monili, sedili, and ab hoc monili, sedili); the reason is that an ablative ending in e would make four cases identical, which the declension of neuter nouns does not permit. Nouns which end with a long e in the nominative, being Greek feminine nouns like Agave and Autonoe, should be declined in the Greek manner so that their nominative, dative, vocative, and ablative are all identical, with the genitive ending in s and the accusative in n. /88/ However, the Greek feminine nouns which change their e ending to a, as Andromache to Andromacha, are declined in the same way as the Latin nouns ending in a.

Latin nouns ending in i are of every gender and are indeclinable, like frugi, nihili. Gummi and synapi, however, are exotic nouns of neuter gender and only singular number; their case endings are identical, except for the genitive, which has an s (synapis, gummis); nevertheless it is correct to say haec gummis.

→ lat [292] No neuter nouns end in o; nouns with this ending are either masculine, like Cicero, or feminine, like Juno, hirundo, or of common gender, like homo. The declension of such nouns takes two forms: either they keep the o in the oblique cases, as in unio, unionis; or they change it to i, as in cupido, cupidinis; crepido, crepidinis; in both instances they become a syllable longer in the oblique cases, except for the vocative singular. Although caro is declined contrary to this analogy, custom has preserved it and Anio also; in their declension they are unique, and follow no exemplars, for their genitives are huius carnis, Anienis. Duo and ambo are always plural, and therefore should not be included amongst the categories I mentioned above; they follow their own particular declensions. Nouns ending in o of Greek origin but later Latinized, such as leo, draco, are declined like Cicero, Milo; those which have kept their earlier form, like Io, Ino, are declined in the Greek way, with their genitives Ius, Inus, accusatives Iun, Inun,⁸² and so on for the other cases, although it is customary to say hanc Io. Turbo as a proper noun is declined like Cicero; but if it means ‘the force of the wind’ or the child’s toy⁸³ it is declined like cupido.

→ lat [293] Only neuter nouns end with the letter u; for instance, cornu, veru, which in the plural is verua. Earlier writers also used specua. The declension of the plural is quite certain in these nouns; the first three cases are the same, as usual in neuter nouns, the dative and the ablative end in bus, and the genitive adds um to the ablative singular. There is dispute about the genitive singular, which some end in us, others in is; also about the dative singular, which some end in i, as in cornui, genui; but some follow archaic writers /89/ and make the dative the same form as the ablative, huic genu, cornu; the grounds for this are, that we say in the plural genubus, cornubus, and when the last syllable is omitted the remainder is the dative singular form, as happens in civibus and suavibus. Just as in genibus and cornibus the letter i has been substituted for the letter u, so do we say optimus and maximus when the forms used to be optumus and maxumus. However, some scholars say genuis and cornuis for the genitive singular, though the genitive ought not have more syllables than the dative and the ablative. Although senatus and exercitus are of a different declension, yet one ought to say genus and cornus for the genitive singular, on the model of senatus and exercitus. Besides, any nouns or participles which end their genitive singular in is have one syllable more in the dative plural, as in Catonis, Catonibus; on this principle, if the genitive singular be genuis, cornuis, the dative plural would be genuibus, to make an extra syllable.

→ lat [294] Nouns ending in l have two forms. The first includes masculine nouns like Hannibal, Hasdrubal, and comprises Carthaginian names almost entirely, except for our name Sol; it also includes feminine nouns like Tanaquil, an Etruscan name, and common nouns like vigil, pugil, which are all of the same declension. The other class is of neuter nouns such as mel, fel, which differ from the first class in being monosyllables and in doubling the l in the oblique cases. Those who say proconsule as a nominative, with the e ending, have in view the nature of the preposition pro, which is used only with an ablative. They do not take the view that when one says proconsul one is using not two words but a composite noun like procurator, propugnator; the fact that these nouns can form verbs like procuro, propugno, where the preposition is absorbed into the verb, while proconsul forms no such verb (though we do have the verb consulo), in no way renders the example invalid. If one should say proconsule as the nominative, then the noun would be indeclinable; but one ought to follow the custom and say proconsul, so that the noun can be declined in the same way as the other nouns ending in l, like vigil, mugil, pugil, where the genitive plural is mugilum and pugilum; for if the nominative were mugilis, like agilis, the genitive plural would be mugilium, like agilium.

→ lat [295] M is the ending of neuter nouns like caelum, scamnum. Note that nouns which have i before the last syllable, like lilium, folium, /90/ ought to double the i in the genitive singular and the dative and ablative plural, as in lilii and liliis; for the genitive has either the same number of syllables as the nominative, as in scamnum, scamni; or more syllables, as in caput, capitis; but never fewer syllables. Besides, the dative ending in o changes the o to i to form the genitive, as scamno to scamni, but usage and the authority of earlier writers prefer to say ingeni and consili and imperi in three syllables.

It is disputed whether one ought to sav vasum or vas, since all neuter nouns whose nominative singular ends in something other than m form their dative and ablative plural in bus; for instance, monile, monilibus; genu, genibus; pecus, pecoribus; nomen, nominibus; marmor, marmoribus. On this principle, vas ought to form vasibus and, in the genitive, horum vasum; for all neuter nouns which end their ablative singular with e have in the genitive plural the same number of syllables as in the ablative singular, for instance, a nemore, nemorum; capite, capitum. However, the declension of vas is an anomaly; in the singular the nominative is vas, genitive vasis, dative vasi, ablative a vase; while in the plural it is declined like scamna. Nevertheless, Lucretius has the genitive vasi, in rarique facit lateramina vasi.⁸⁴

→ lat [296] N is the final letter of masculine nouns like flamen, pecten, and of neuter nouns like omen, flumen, and of nouns common to both masculine and feminine genders, like tibicen, fidicen --- though some authors disagree with this, using tibicinam and fidicinam.⁸⁵ All these nouns follow the same clearly defined declension. Gluten is similar to these, though Sallust uses glutinum. Some feminine nouns seem similar, such as Siren. There is another kind of masculine noun, such as lien, rien, which keeps the e before the final n, but in other respects declines just like the examples cited. Ren is a nominative singular which has no other case in use except the ablative. Greek nouns ending in n have before the n the letters a, e, i, o, as in Alcman, Cephen, delphin, Phaethon; of these, those which end in an, en, in follow the declension of the examples above, like fulmen, numen, fidicen; those which end in on and in Greek retain the long o throughout their declension, such as leon, keep the o long in Latin also and are declined like Cicero, Scipio; those which in /91/ Greek change the long o of the nominative to a short o in the other cases, such as Amphion, Creon, Agamemnon, in Latin lose the r in the nominative and are declined like those Latin nouns which shorten the final vowel of their stems in the oblique cases, such as virgo, turbo. Those which in Greek add t to the nominative do the same in Latin, as Phaethon, Phaethontis.

→ lat [297] Nouns which end in ar, like Caesar, lar, far, par, impar, follow a single formation, except that neuter nouns double the r in the oblique cases, as far, farris [grain]. Poetic license uses a plural with these nouns, although there is no plural for those nouns whose quantity we measure by weight or extent, like aurum [gold], plumbum [lead], triticum [wheat], oleum [oil]....⁸⁶ So people are mistaken when they say parium, because these nouns have the same ending in the first three cases of the plural, as in hi Caesares, hos Caesares, o Caesares, and the accusative cannot have i before the final s if the genitive has r before the u of its ending.⁸⁷ But those who say parium appear to be considering the nominative plural of the neuter;⁸⁸ and since the nominative plural is paria, like suavia, they say for the genitive plural parium, like suavium.

→ lat [298] There are six variations on the ending er: the first when the genitive has the same number of syllables as the nominative and ends in i, as aper, niger, macer, giving apri, nigri, macri; the second like tener, lacer, puer, giving teneri, laceri, pueri; the third like imber, liter, which is to this extent different from the foregoing, that the genitive singular ends in s and the dative in i, yet they are not like the second type longer than the nominative, and the genitive plural ends in ium; the fourth type is like mater, pater, frater, which form their dative plural not in is (apris) like the first type but in ibus (fratribus), while at the same time this type does not lengthen by a syllable in the oblique cases like the second type, and has the genitive plural ending um; for we say patrum, fratrum, not (as in the third type) utrium, imbrium. Besides, this fourth type /92/ has the same ending in the three cases hi patres, hos patres, o patres, while imber and liter have hos inibris and utris. The fifth type is that of passer, anser, later, mlier, which differs from the second type in that the second type forms its genitive ending in i, and it is longer by one syllable than the nominative (as in gener, generi), while the fifth type has the longer genitive but its ending is in is (as in passer, passeris). There are other differences as well. The natural classification of piper and cicer is in this fifth type, for they are declined like passer in the singular, and they occur only in the singular. The sixth type is that of neuter and uter; it forms neutrius and utrius, and its declension is that of the pronouns which have the genitive ending in ius and the dative in i; these genitive and dative endings are common to all genders.

There is only one type with the nominative ending ir, that of vir. There are two types with the ending or; one lengthens the o in the oblique cases, as in sopor, soporis; color, coloris; the other has the o short, as in arbor, arboris; they are declined in the same way. In former times the genitives Castoris and Hectoris were pronounced with the o long, but nowadays we make them short, since the Greeks pronounce them short. There are two types with the nominative ending nr; the first, like satur, has the genitive ending i, as in saturi; while the second has the genitive ending is, as in sulphur, sulphuris; augur, auguris. Of the neuter nouns with the nominative ending ur, some keep the u, as in sulphur, sulphuris; others lose it as in ebur, eboris.

→ lat [299] There are eight types with the nominative ending in s; one type has a before s, like Maecenas, civitas; one has e, like verres, moles; or i, like panis; or o, like custos, nepos; or u, like vetus, Ligus; or r, like iners; or n, like serpens; or p, like praeceps. The type with a before s has two forms: The first like Maecenas, Laenas; the second like nostras, Privernas, which differ from the first form in that they are common to all genders and ought to have in the genitive plural the letter i, which is frequent in the nominative, accusative, and vocative plural of neuter nouns. Thus we say nostratium, Privernatium, since the nominatives are nostratia, Privernatia; we also say praegnatium, optimatium, because these have their accusative plural in i, optimatis, praegnatis, although the older writers used praegnatum, optimatum; praegnas can be both feminine /93/ and neuter in gender. As and mas, being monosyllables, are not bound by the analogy, but follow their own forms of declension, giving assis and maris for the genitive singular, and assium and marium for the genitive plural. Fas and nefas are indeclinable.

→ lat [300] There are three kinds of Greek nouns which, when used in Latin, end in as: the first, like Olympias, Pythias, form their genitive Olympiadis, Pythiadis; the second, like Pallas, Thoas, Atlas, form it Pallantis, Thoantis, Atlantis; the third, like Aeneas, Pythagoras, Lichas, form it Aeneae, Pythagorae, Lichae, following the declension of those Latin nouns which have their genitive in ae and their nominative in a, like Catilinae from Catilina. These nouns, being Greek, keep the s in the nominative, although some Greek nouns, like Nicia, Mela, lose it in the nominative; in these nominatives one ought to follow customary usage.

→ lat [301] There are five kinds of nouns ending in es: the first have their nominative, vocative, and accusative plural the same, but the genitive plural ends in um, as Hercules, proles, give Herculum, prolum. The second kind differs from the first in that the genitive plural ends in ium and the accusative plural in is as with nubes, rupes, cautes; there are no masculine nouns in this class. The third kind might have some differences of declension amongst themselves but are considered one class because the nouns agree in adding an extra syllable in the oblique cases; such nouns are Ceres, bipes, merces; there are no masculine nouns in this class except Greek nouns like Chremes, Laches. The fourth kind differs from the third in changing e into i in the oblique cases, whereas the third kind retains the e; and whereas the third kind has the e long in the nominative singular, the fourth kind has it short, as in hospes, antistes, ales, comes. The fifth kind differs from the fourth in that the genitive has an extra syllable and ends in i, which is the same as the dative, while the accusative ends in em and the ablative in long e; the addition of rum to this makes the genitive plural. The nominative plural is identical with the nominative and vocative singular, while the dative and ablative plural ends in the syllable bus, as in facies, dies, spes, acies. Rei and spei are customarily pronounced short, perhaps because these nouns are monosyllabic.

→ lat [302] Of the nouns ending in is, some add a syllable in the oblique cases, others do not, at least in the singular. There are two kinds of /94/ these latter; the first includes masculine nouns, like scrobis, mensis (although Lucan says exigua scrobe),⁸⁹ feminine nouns,⁹⁰ and nouns of common gender like canis, iuvenis, the declension of which is clear. The second kind differs from the first in that the ablative ends in i, the accusative in im, as in the only two masculine examples, Ligeris and Tiberis, and the feminine ones, clavis, pelvis, turris, sitis, tussis. Their diminutives show clearly that in their declension all these are pronounced with a long i. So as often as we are in doubt over the pronunciation of a noun, let us look at its diminutive. Of those which add a syllable there are two kinds: one keeps the i short, and this includes masculine nouns like sanguis, pulvis, lapis, and feminine nouns like cuspis, cassis; it makes no difference that in their declension one group keeps the i while the other converts it to e (in the ablative), nor that one ends its genitive singular in dis, the other in ris. The second kind has the letter i long in the nominative, as in glis, lis, Samnis, and keeps it so in the oblique cases, while in the genitive plural we use i before the final um, as in glirium, litiumm, Samnitium, as distinct from the first kind’s lapidum, cuspidum, cassidum.

→ lat [303] Nouns ending in os can be said to form one class, since there is little difference between them. Though they all lengthen by a syllable in the oblique cases, they differ in that some convert the s of the nominative to t, some to d, some to r, as seen in nepos, nepotis; custos, custodis; masculine monosyllables like flos, floris, ros, roris, or a neuter one like os, oris. In ossibus, however, the nominative is os by usage and contrary to principle, since it does not have r in the genitive; but the nominative ossum is impossible, since neuter nouns with the nominative singular ending in m do not admit the syllable bus in the plural; so in the monosyllable the analogy does not apply.

→ lat [304] There are six kinds of nouns ending in us. In the first the genitive ends in the single i, the vocative in e: for instance, the genitive of Marcus Sextus is Marci Sexti, the vocative o Marce Sexte. The second kind differs from the first in that the genitive ends in double i and the vocative ends in i: for instance, Antonius Iulius gives the genitive buius Antonii Iulii and the vocative o Antoni Iuli; otherwise the second kind is the same as the first. The third kind had /95/ the same form in the nominative and the vocative, and the genitive is spelled the same but pronounced with a long ending; the dative ends in ui, the ablative in u, the accusative in um, as with senatus, fluctus, exercitus; in the plural these nouns have long us in the nominative, the accusative, and the vocative; the genitive ends in uum; the dative and the ablative in bus. The fourth kind adds a syllable in the oblique cases, but the nouns have minor variations amongst themselves; some keep the letter u, pronounced short; others have it long, as in Ligus, Liguris; palus, paludis; virtus, virtutis. Some change the u to e, as vetus, veteris; Venus, Veneris; some to o, as nemus, nemoris; but these follow the same declension. Although laus and fraus are monosyllables containing diphthongs, they are in this same class, because they also add a syllable in the oblique cases. The fifth kind is the class of pronouns which end in us, such as unus, solus, totus; they are declined thus: unus, unius, uni, unum, unus, ab uno, uni, unorum, unis, unos, uni, ab unis. The sixth kind ends in eus, like hinnuleus, eculeus, and in the vocative the e is repeated, as in eculee, hinnulee; though some people prefer to use the Greek vocative in eu, as in Tydeu.

→ lat [305] Neuter nouns ending in is are indeclinable, like tressis, sexis. There are two kinds of neuter nouns ending in us. The first has the genitive in i and has no plural: for example, vulgus, pelagus. Lucretius says viri from virus,⁹¹ although it would be more correctly left uninflected. The second kind includes those nouns which add a syllable in the oblique cases and end the genitive singular in ris, such as genus, nemus; some of these change the u of their nominative ending to e, as olus, oleris; ulcus, ulceris; some change it to o, as nemus, nemoris; pcus, pecoris. There is some doubt whether fenus and stercus change their u to e or to o, since nouns like vulnus, genus, funus, ending in the syllable nus, change the u to e; and we do use the word feneratus. The example of penus does not affect the issue, since its place is amongst the doubtful kinds of noun. Again, earlier writers used stercoratos agros, not sterceratos.

→ lat [306] Nouns ending in ns or rs are all of one kind, except that of those ending in rs some end their stem in the oblique cases in d, like socors, socordis; some in t, such as sollers, iners. Their declension is the same in all cases of the plural except the genitive and the accusative; /96/ some, like Mars and Arruns, end the genitive plural in um and the accusative in es; some, like sapiens and patiens, end the genitive plural in ium and therefore end the accusative in is. Most nouns of this group are common to all three genders, and the letter i, which appears in the nominative plural neuter, is found also in the genitive of the other genders with which the noun is common.

→ lat [307] Only a few neuter nouns end with the letter t, such as the indeclinable git, and caput and sinciput. Some people, when they say lac, add t, because it does form the genitive lactis; but Vergil writes lac mihi non aestate novum non frigore defit;⁹² indeed no nouns in Latin end in two mutes, and for this reason archaic writers used lacte as the nominative.

→ lat [308] Of the nouns which end in the letter x, some (amongst which are all the common nouns) end their genitive plural in him, and therefore have their accusative in is; but most end their genitive plural in um without the antecedent i, and therefore have their accusative in es. In all other details both kinds agree; in the singular, all end the nominative in x, the genitive in is, the dative in i, the ablative in e, and add to this m to form the complete accusative; in the plural the dative and ablative end in the syllable bus. Oh the points where they differ, some earlier writers used astrocum and ferocum; if one follows them, there appears to be only one class of nouns ending in x. Any vowel may precede the x, as in capax, frutex, pernix, atrox, redux. Some of these nouns pronounce the last syllable of the nominative long, some short; some nouns follow the same practice as each other in the nominative, but differ in the oblique cases. Pax and rapax, rex and pumex, nux and lux differ in pronunciation of the nominative; but nix and nutrix, nox and atrox give the same value to their corresponding vowels in the nominative, but different values in the oblique cases.⁹³ We should note that in the declension of these nouns some change the x to g, some to c: for lex makes legis; grex, gregis; but pix makes picis; and nux, nucis. The x always changes to c in the genitive of nouns that are not monosyllabic, as in frutex, fruticis; ferox, ferocis. By exception, supellex, senex, and /97/ nix are declined contrary to the general rule: supellex adds two syllables to its stem in the oblique cases, which is against the rule; senex remains disyllabic in the genitive as in the nominative, although all other nouns ending in x add a syllable; and nix changes its stem ending not to c, like pix, nor g, like rex, but to v, although a consonant may not change to a vowel.

The form of the genitive plural is derived from the ablative singular. If the ablative singular ends in a or o, the ending of the genitive plural adds rum to that; if the ablative singular ends in short e, the genitive plural ending is um; if the ablative singular ends in long e, the genitive plural ending is rum; if the ablative singular ends in i, the genitive plural ending is ium; if the ablative singular ends in u, the genitive plural ending is uum. The dative and ablative plural end either in is or in bus --- these are well-worn school precepts; when they end in is the syllable is long, when in bus it is short.

→ lat [309] After that survey of the rules for nouns, it is proper to come on to those for verbs. There are five kinds of verbs: active, passive, neuter, common, and deponent. The active end in o and signify action, such as lego [I read], scribo [I write]; the passive end in r and signify that one is acted upon, as with scribor [I am written], legor [lam read]; the neuter [intransitive] end in o and do not signify complete action or complete passivity; such are sudo [I sweat], dormio [I sleep]; in these, I do not know whether someone is acting or being acted upon. Common and deponent verbs both end in r, but the difference between them is this: the common verb may have both active and passive meanings, for we say osculor te [I kiss you] and osculor a te [I am kissed by you]; but the deponent verb has either an active meaning, as in luctor [I grieve], or a passive, as in morior [I die]. There is also the impersonal verb, such as sudatur, curritur,⁹⁴ which is so called because although it may apply to any person it has no particular person as its subject.

→ lat [310] Verbs have five moods, although some people have said they have six, some seven, some eight, some nine, and a few ten. Those who say five name these: indicative, imperative, optative, conjunctive, infinitive --- which is also called constant. Those who say six /98/ add the permissive, those who say seven add the impersonal, those who sav eight add the interrogative, those who say nine add the subjunctive and distinguish it from the conjunctive, those who say ten even add the hortative; but these unnecessary additions are against all reason.

→ lat [311] There is no disputing that there are three conjugations (syzygiae, as the Greeks call them), which in active and neuter verbs are distinguished in the second person of the present tense. As often as the last syllable of this ends in as, it is the first conjugation; in es, the second conjugation; in is, the third: for example, cantas, vides, audis. Audis would exemplify the ‘third conjugation, long,’ while the ‘third conjugation, short’ would be as in curris.⁹⁵ Verbs of all these conjugations, in the first person singular of the present indicative active, permit only three vowels to precede the last letter; namely, e, i, u, as in sedeo, lanio, irruo; but any consonant may precede except f, k, and q: for example, libo, voco, cado, lego, trabo, impello, avio, cano, scalpo, curro, lasso, peto, nexo. To these are added i and v when used as consonants, as in aio, adiuvo. F is not excluded when we say triumfo, though this comes from Greek and is better written with ph; besides this example, from the fact that we say faris and fatur people postulate a first person of the verb, for.⁹⁶ There is also some doubt about the letter q, for we do say eliquo and aequo, and in words of this kind the letter u has the value of neither a vowel as in irruo nor a consonant as in adiuuo.

→ lat [312] In declining all these conjugations there are thirty-six forms apart from the defective, the impersonal, and the inchoative. Verbs of the first conjugation, whether they end with no other vowel preceding the o or with any vowel preceding it, have four forms. Verbs of the second conjugation have seven forms. Verbs of the ‘third conjugation, short’ have twenty forms; those which end their first person singular present indicative with io have six forms; those that end this part with uo have two forms; those that end this part with o with no vowel preceding have twelve forms. Verbs of the ‘third conjugation, long,’ which end their first person singular present /99/ indicative with io, have five forms. Any verbs of any conjugation whose first person singular present indicative ends with o with any other vowel preceding or with no vowel preceding has its declension included in this number of forms, which I shall discuss one by one.

→ lat [313] Verbs of the first conjugation end the first person singular present indicative with o, either with no other vowel preceding, such as amo, canto; or with eo, such as commeo, calceo; or with io, such as lanio, satio; or with uo, such as aestuo, continuo. Verbs of the first conjugation end the second person present imperative with long a, as ama from amo, canta from canto. They end the infinitive by adding re to the imperative, keeping the vowel long, as in amare from ama, cantare from canta. Again, in the first conjugation, verbs which end their perfect indicative by adding vi to the imperative, keeping the syllable long (as in commeo, commea, commeavi, and lanio, lania, laniavi, and satio, satia, satiavi), end the imperfect by adding bam to the imperative, as in commeabam from commea, laniabam from lania, aestuabam from aestua. The first conjugation ends its pluperfect by adding veram to the imperative, as in commeaveram from commea, laniaveram from lania, aestuaveram from aestua. In this conjugation bo added to the imperative ends the future indicative, as in commeabo from commea, laniabo from lania, aestuabo from aestua.

→ lat [314] Verbs which end their first person singular present indicative with o with no preceding vowel appear in the perfect and pluperfect indicative in four different ways. The rule of the first is similar to the preceding verbs which in the first person singular present indicative have a vowel immediately before the last letter of the stem, as in amo, ama, amavi, amabam, amaveram, amabo, amare. The second way changes the o to i and lengthens the penultimate vowel in the perfect and the antepenultimate vowel in the pluperfect, as in adiuvo, adiuvi, adinveram. The third way follows a rule similar to the first but separates and omits the a, as in seco, seca, secui, secabam, secueram, secabo, secare; it makes secui in the perfect and secueram in the pluperfect. The fourth way appears through reduplication of the syllable, as in sto, sta, steti, stabam, steteram, stabo, stare; similar to this is do, da, dedi, dabam, dederam, dabo, dare, where the letter a is shortened, contrary to the rule applying in the formation of dabam, dabo, dare.

/100/ → lat [315] Verbs of the second conjugation end the first person of the present indicative with eo, as in video, vides; moneo, mones. They end the present imperative second person with long e, as in video, vide; moneo, mone. The infinitive is formed by adding re to the imperative, keeping the syllable long, as in vide, videre; mone, monere. The perfect and pluperfect indicatives are conjugated in seven different ways. The first is the regular form: the perfect is formed by adding to the imperative ending (keeping the syllable long) the letters vi, as in deleo, dele, delevi; the pluperfect is formed by adding veram to the imperative ending, as in dele, deleveram. The second way is by changing the eo ending of the first person present indicative to i, as in sedeo, sedi; and for the pluperfect, sederam. The third way is when eo changes to ui, as in caleo, calui; moneo, monui; calueram, monueram. The fourth way is when exchanges to i and the consonant preceding eo changes to s, as in mulgeo, mulsi, mulseram. The fifth way is when eo changes to i and the consonant preceding eo changes to x, as where luceo and lugeo both form luxi, luxeram. The sixth way is when the part is formed by reduplication, as in spondeo, spopondi, spoponderam. The seventh way is when the perfect tenses are of passive form, as in video, ausus sum, es, est; ausus eram, eras, erat. Verbs of the second conjugation form the imperfect indicative by adding bam to the imperative ending, keeping the syllable long, as in vide, videbam; mone, monebam. They form the future by adding bo to the imperative ending, keeping the syllable long, as in vide, videbo; mone, monebo.

→ lat [316] Verbs of the third conjugation, short, in the first person present indicative either end in the letter o with no other vowel preceding it, as in lego, legis; peto, petis; or end in io, as in rapio, rapis; facio, facis; or in uo, as in induo, irruo. Verbs of this conjugation, which end the first person singular present indicative in o with no preceding vowel, end the second person singular present imperative in short e with the same consonant preceding the ending as in the present indicative, as in lego, lege; peto, pete; those verbs of this conjugation which end in uo retain the u with the e of the imperative, as in induo, indue; irruo, irrue. Verbs of this conjugation end their infinitive by adding re to the imperative, keeping the e of the imperative short, as in lege, legere; pete, petere. → lat [317] Those verbs of this conjugation which end their first person present indicative in io form their perfect and pluperfect tenses in one of six ways. /101/ The first way is to follow the rule of the third conjugation, long, making the i of the imperative long and adding for the perfect the syllable vi, for the pluperfect veram, as in cupio, cupivi, cupiveram. Sometimes the v is omitted and in the perfect the i is repeated (cupii), while in the pluperfect it is pronounced short (cupieram). The second way is to omit the o of the first person singular present indicative and to lengthen the preceding syllable, either altering the vowel or leaving it, as in facio, feci; fugio, fugi; in the pluperfect the i is replaced by e, and the syllable ram is added, as in feceram, fugeram. The third way is when the io ending is turned into the ending ui, as in elicio, elicui, elicueram; the fourth way is by dropping the io ending and the preceding consonant and pronouncing the word with a double s, as in percutio, percussi, percusseram; the fifth way uses x instead of ss, as in aspicio, aspexi, aspexeram; the sixth way reduplicates a syllable, as in pario, peperi, peperam. → lat [318] Verbs of this conjugation which end the first person singular present indicative in no form their perfect and pluperfect in two ways. The first is by changing the o of the present indicative to i, as in induo, indui, indueram; the second uses x, as in instruo, instruxi, instruxeram. → lat [319] Verbs of this conjugation which end their first person singular present indicative in o with no preceding vowel form their perfect and pluperfect indicative in twelve ways. The first is the same as the third conjugation, long, as I showed above in dealing with those verbs of the third conjugation, short, whose first person present indicative ends in io; for as cupio forms cupivi and upiveram, so peto forms petivi and petiveram. Sometimes also the v is omitted and the i repeated, as in petii, petieram. The second way is to change the o of the present indicative to i...,⁹⁷ to change the o of the present indicative to i and lengthen the preceding syllable, either altering its vowel or preserving it, omitting also any penultimate consonant which closed the syllable in its original form; for instance, ago, egi, egeram; lego, legi, legeram; or, as instances of those which omit consonants, frango, fregi, fregeram; fundo, fudi, fuderam. The fourth way is to omit the penultimate consonant of the present stem and shorten the preceding vowel, as in findo, fidi, fideram; scindo, scidi, scideram. The fifth way is to substitute ui for the o of the present indicative, as in molo, molui, molueram; colo, /102/ colui, colueram. The sixth way is to substitute si for the o, as in carpo, carpsi, carpseram; scribo, scripsi, scripseram. The seventh way is to omit the o and turn the preceding consonant into double as in vieto, messui, messueram. The eighth has a similar omission, of o, a single s, and the preceding vowel long, as in trudo, trusi, ruseram. The ninth uses x, as in expungo, expunxi, expunxeram; ungo, unxi, unxeram. The tenth reduplicates the first syllable, as in pungo, pupugi, pupugeram; curro, cucurri, cucurreram. The eleventh way reduplicates the final syllable of composite verbs, as in trado, tradidi, tradideram; reddo, reddidi, reddideram. The twelfth way uses a passive form, as in fido; fisus sum, es, est; fisus eram, eras, erat. → lat [320] All these verbs of the third conjugation, short, end their imperfect indicative by adding the syllable bam to the imperative, with the same consonant or vowel to end the stem as in the first person singular present indicative; for instance, lege, legebam; pete, petebam. The exceptions are those verbs which have i before the o of the present indicative; for they form their imperfect ending by inserting i before the imperative ending, lengthening the e and adding bam, as in rapio, rape, rapiebam. Verbs of this conjugation form the ending of the future indicative by dropping the e of the imperative form and adding the syllable am, as in lege, legam; pete, petam; indue, induam. The exceptions are those verbs which have i before the o of the present indicative; to form the future they change the e of the imperative to i and add the syllable am, as in rapio, rape, rapiam; facio, face faciam.

→ lat [321] Verbs of the third conjugation, long, in the first person singular present indicative end either in eo, as in adeo, adis; prodeo, prodis; or in io, as in audio, audis; nutrio, nutris. Verbs of this conjugation in the second person singular present imperative end in long i, as in adeo, adi; prodeo, prodi; audio, audi; nutrio, nutri. In the infinitive mood they end by adding re to the imperative, keeping the i long, as in adi, adire; prodi, prodire. Verbs of this conjugation which end their first person present indicative in eo end their perfect by adding vi to the imperative, as in adeo, adi, adivi, and the pluperfect by adding veram, as in adiveram; but usage for the sake of brevity frequently omits the v and in the perfect reduplicates the i, as in adii, adieram; prodii, prodieram. The ending of the imperfect is formed by adding bam to the imperative, as in adibam, prodibam.

/103/ → lat [322] However, those verbs of this conjugation which end their first person present indicative in io form their perfect and pluperfect in five ways. The first is like that mentioned above: nutrio, nutrivi, nutriveram, nutrii, nutrieram; and audii, audieram, as when Vergil says audieras et fama fuit.⁹⁸ Only the imperfect form differs from the example above; for all the verbs of this conjugation which end their present indicative first person in io form their imperfect in the same way, by adding long e before the final syllable, as in audiebam, veniebam, operiebam; although earlier writers used these without the letter e, as when Vergil says: nutribat Tyrrhusque pater, cui regia parent / armenta.⁹⁹ The second way is to omit the u of the present and to lengthen the preceding syllable, as in venio, veni, veneram. The third way is to substitute ui for io, as in operio, operui, operueram. The fourth way is to omit io and change the preceding consonant to s, as in sarcio, sarsi, sarseram. The fifth uses x, as in vincio, vinxi, vinxeram.

→ lat [323] Verbs of this conjugation which end their present indicative in eo form their future ending in the one way, by adding bo to the imperative, as in adeo, adi, adibo; prodeo, prodi, prodibo. Those verbs which end their present indicative in io add am to the imperative to form the future, as in audio, audi, audiam; nutrio, nutri, nutriam. Terence, however, uses the bo ending, so that there appear to be two alternative formations; for he says: iam scibo, ubi siet.¹⁰⁰

→ lat [324] Of those verbs which have a present indicative ending in eo there are some which, though differing in their present forms, have an identical perfect form, as for instance luceo and lugeo: for the perfect of both is luxi. Similarly, cernit and crescit have the perfect form crevit. There are other verbs which, though sharing a common first person singular present indicative, are of two different conjugations, such as pando and mando; for they form both pandas and pandis, both mandas and mandis.

Verbs sometimes govern a single case, like the genitive as with misereor tui, or the dative as with suadeo tibi, or the accusative as with moneo te, or the ablative as with utor illo; sometimes they /104/ govern two cases, like the genitive and...¹⁰¹ the accusative and ablative as with fungor hanc rem and fungor illa re.

Those are enough examples of analogy. The forms that are not included above must certainly be reckoned as anomalies. I shall discuss these very briefly to show that some words have acquired a usage inconsistent with their form or have abandoned a usage proper to it.

→ lat [325] When reus and deus are similar in the nominative singular, why do we say hi rei in the plural for one and, inconsistently, hi di for the other, when we ought to say dei --- especially when no genitive should be longer by two syllables than its nominative, as happens when we sav deorum? When Thoas, Aeas, and Aeneas are similar, why do two form the genitives Thoantis and Aeantis, yet Aeneas forms not Aeneantis but Aeneae? When we say hic biceps and triceps, why is the genitive form two syllables longer, contrary to the norm, so that we say bicipitis or tricipitis, and not bicipis or tricipis? How does it happen that aliger, frugifer, accipiter have all the case terminations, but Iuppiter has only two? When sanctus, pius, and bonus are all similar, why do we say sanctior but not piior? Why do we say sanctior, sanctissintus, but not bonior, bonissimus? When Vergil says fandi atque nefandi,¹⁰² why do we form nefarius from nefando but not farius from fando? Why can seiunctus lose its prefix and form iunctus, while if you take se away from securus and sedulus you cannot use what is left? When we may say, in the plural, singuli viri, singulae mulieres, singula scrinia, why may we not say, in the singular singulus vir, singula mulier, singulum scrinium? When the verbs venor, piscor, and aucupor are similar, why do we form from them the nouns Venator and piscator, yet not aucupator but auceps? Why does volo have no imperative? Why does the verb fare have no first person singular in the present indicative? Why does soleo have no perfect tense? When canta and lava are similar, why does one form cantavi but the other not form lavavi? Again, corusca forms coruscavi, but tona does not form tonavi. Why does ego have no other case? /105/ When calceatus, armatus, togatus, and penulatus all seem the same, why do we say calceo and armo and cannot say togo and penulo? Adjectives whose nominative singular ends in us all form the adverb with a long e ending, as doctus forms docte; avarus, avare; parens, parce; then why do bonus and malus have the e of the adverb short in bene and male? When we form the adverb habiliter from habilis, why do we not form faciliter from facilis? Again, when we say difficulter, why can we not say faculter? When from audax we form audacter, why do we form from verax not veracter but veraciter? Why do we say singulatim but not binatim or ternatim? There are innumerable things of this kind which I could mention if I did not have to hurry on to other topics.”

→ lat [326] When Grammar had said this as if she were merely introducing her subject, Minerva intervened, because of the boredom that had come upon Jove and the celestial senate, and said: “Unless I am mistaken, you are getting ready to go back to the elements and begin telling us about the eight fundamental parts of speech, adding also the causes of solecisms, the barbarisms, and the other faults of speech which celebrated poets have discussed at length; you will also discuss tropes, metaplasms, schemata, figures, and all the faults which flow, as it were, from the fountain of embellishment, illustrating either the misconception of the writer who docs not understand them or the labored ornamentation of the pedant. If you bring such matters from the elementary school before the celestial senate, you will nip in the bud the goodwill you have won by this display of knowledge. If you were to take up a discussion of rhythm and meter, as you would venture to do with young pupils,¹⁰³ Music would surely tear you apart for usurping her office. The teaching you have given us will be well-proportioned and complete if you keep to your own particular subjects and do not cheapen them by commonplace and elementary instruction.”

The Delian and his spouse nodded in approval of these words of his sister, and they made Grammar walk across to the attendants of Philology. Then the Clarian introduced another of the women who would form part of the nuptial exchange of gifts.