# How to Scan Dactylic Hexameter, Step by Step

Learn to scan dactylic hexameter step by step: long and short syllables, elision, the six feet and the caesura, with Aeneid 1.1 and Iliad 1.1.

Source: https://lemma.algon.one/en/blog/how-to-scan-dactylic-hexameter · Published 2026-10-13 · Topics: latin, greek, metre, poetry

To scan a line of verse is to mark, syllable by syllable, which syllables are **long** and which are **short**, and then to group them into feet. Greek and Latin poetry is built on syllable length (**quantity**), not on stress as English verse is. That is why scansion feels strange at first, and also why it behaves like a puzzle with exact rules. This guide gives a method for the **dactylic hexameter**, the metre of Homer, Virgil, Lucretius, Ovid's *Metamorphoses* and Horace's *Satires*.

## The pattern: six feet, a fixed ending

A hexameter has **six feet**. Each foot is either a **dactyl** (one long and two shorts: — ∪ ∪) or a **spondee** (two longs: — —). Two facts make the end of the line predictable:

- the **fifth foot** is nearly always a dactyl;
- the **sixth foot** has only two syllables: a long, then a final syllable that may be long or short. That last syllable is called *anceps* and is marked ×, so the foot is a spondee or a trochee (— ∪) in practice.

Feet one to four may each be a dactyl or a spondee. The full pattern is:

> — ∪ ∪ | — ∪ ∪ | — ∪ ∪ | — ∪ ∪ | — ∪ ∪ | — ×

where any `— ∪ ∪` in feet one to four may be replaced by `— —`.

## When is a syllable long?

A Latin syllable is **long** in two cases.

1. **Long by nature**: it contains a long vowel (*ā, ē, ī, ō, ū*) or a diphthong (*ae, au, oe*, and a few others, such as *eu* in *heu* and *seu*).
2. **Long by position**: it contains a short vowel followed by **two consonants**, or by a double consonant (*x* or *z*). The two consonants may sit in the same word or be split between two words.

A syllable that fits neither case is **short**. Four details prevent most mistakes:

| Detail | Rule |
|---|---|
| *h* | does not count as a consonant |
| *qu* | counts as a single consonant |
| consonantal *i* | at the start of a word (*iam*, *iuvenis*) it is one consonant; between two vowels (*maior*, *eius*, *Troia*) it normally counts double and makes the syllable before it long (*Troi-ia*) |
| stop + liquid | when a stop (*b, c, d, g, p, t*) is followed by *l* or *r* in the same word, a short vowel before the pair **may** stay short; the poet chooses |

Greek works the same way. The vowels η and ω are long by nature, ε and ο short, while α, ι and υ may be either; diphthongs are long; ζ, ξ and ψ count as double consonants.

Vowel length is not shown in ordinary printed texts, so you often need to know the form: the ablative *sonitū* ends in a long vowel, the first-person ending of *canō* is long. A dictionary that marks quantities helps, and so does a tool such as Lemma, which shows the form and the headword of every word in a passage and names its source. Careful parsing, as described in [how to translate Latin step by step](/en/blog/how-to-translate-latin-step-by-step), pays off twice in verse.

## Elision

In Latin, when a word ends in a **vowel** (or a vowel + *m*, or a diphthong such as *-ae*) and the next word begins with a vowel or with *h*, the final vowel (together with the *m*) is **elided**: it does not count in the line. Virgil gives two elisions in one verse:

> *litora, multum ille et terris iactatus et alto* (Virgil, *Aeneid* 1.3)
> "…shores, much tossed about, that man, both on land and on the deep"

Read *multum ille* as *mult' ille* (the *-um* drops) and *ille et* as *ill' et* (the *-e* drops). If you forget the elisions, the line has two syllables too many and nothing fits. In Greek there is nothing to strike out: the printed text already marks each elision with an apostrophe.

## The method in six steps

1. **Read the line aloud** and divide it into syllables, treating each diphthong as one syllable.
2. **Mark the elisions** (Latin): strike out a final vowel (or vowel + *m*) before a word beginning with a vowel or *h*. In Greek the elisions are already printed with an apostrophe.
3. **Mark the longs you can be sure of**: diphthongs, vowels you know to be long by nature, short vowels before two consonants.
4. **Start from the end**: the last syllable is anceps and the one before it is long; the fifth foot is nearly always a dactyl. So the last five syllables are almost always `— ∪ ∪ — ×`.
5. **Now work from the beginning**: every foot starts with a long. If two shorts follow, the foot is a dactyl; if another long follows, it is a spondee.
6. **Check**: the six feet must use up every syllable, none left over and none missing. If they do not, look again at elisions and consonant clusters.

## Worked example: a line that gallops

> *Quadrupedante putrem sonitu quatit ungula campum* (Virgil, *Aeneid* 8.596)
> "The hoof shakes the crumbling plain with a four-footed beat"

Seventeen syllables, the maximum: five dactyls and the closing foot.

| Foot | 1 | 2 | 3 | 4 | 5 | 6 |
|---|---|---|---|---|---|---|
| Syllables | Quad-ru-pe- | -dan-te pu- | -trem so-ni- | -tu qua-tit | un-gu-la | cam-pum |
| Quantities | — ∪ ∪ | — ∪ ∪ | — ∪ ∪ | — ∪ ∪ | — ∪ ∪ | — × |

Why is each foot's first syllable long? *dan* is long by position (*n* + *t*), *trem* by position (*m* + the *s* of the next word), *un* by position (*n* + *g*), *cam* by position (*m* + *p*); *tu* is long by nature (ablative *sonitū*). The line also shows the stop + liquid rule both ways: the syllable before *dr* in *Quadru-* counts as long, while the one before *tr* in *putrem* stays short.

## Worked example: Aeneid 1.1

> *Arma virumque cano, Troiae qui primus ab oris* (Virgil, *Aeneid* 1.1)
> "I sing of arms and of the man who first from the shores of Troy…"

Fifteen syllables: dactyl, dactyl, spondee, spondee, dactyl, spondee.

| Foot | 1 | 2 | 3 | 4 | 5 | 6 |
|---|---|---|---|---|---|---|
| Syllables | Ar-ma vi- | -rum-que ca- | -no Troi- | -iae qui | pri-mus ab | o-ris |
| Quantities | — ∪ ∪ | — ∪ ∪ | — — | — — | — ∪ ∪ | — × |

*Ar* is long by position (*r* + *m*), *rum* by position (*m* + *qu*). *no* is long by nature (*canō*); the *Tr* that follows is a stop + liquid and would not have lengthened it. *Troi* is long because the *i* of *Troia* stands between two vowels and counts double (*Troi-iae*); *iae* has the diphthong *ae*; *qui* is long by nature. The fifth foot is the expected dactyl: *pri* long by nature, *mus* and *ab* short.

## Worked example: Iliad 1.1

> μῆνιν ἄειδε θεὰ Πηληϊάδεω Ἀχιλῆος (Homer, *Iliad* 1.1)
> "Sing, goddess, of the wrath of Achilles, son of Peleus"

| Foot | 1 | 2 | 3 | 4 | 5 | 6 |
|---|---|---|---|---|---|---|
| Syllables | μῆ-νιν ἄ- | -ει-δε θε- | -ὰ Πη- | -λη-ϊ-ά- | -δεω Ἀ-χι- | -λῆ-ος |
| Quantities | — ∪ ∪ | — ∪ ∪ | — — | — ∪ ∪ | — ∪ ∪ | — × |

Two things need a comment. The final α of θεά is long, which gives the spondee in the third foot. And the ending -εω of Πηληϊάδεω is pronounced as **one** long syllable, a fusion of two vowels called *synizesis*; without it the line would have a syllable too many. The last syllable, -ος, is short: here the sixth foot is a trochee in practice, which the anceps allows.

## The caesura

A line is more than a row of feet. It also has an internal pause, the **caesura**: a place where a **word ends inside a foot**. The commonest one in Latin falls after the first long of the **third foot** (the penthemimeral caesura, a "strong" caesura in the third foot):

> Arma virumque cano, ‖ Troiae qui primus ab oris
> μῆνιν ἄειδε θεὰ ‖ Πηληϊάδεω Ἀχιλῆος

In the galloping line the break comes in the same place, after *putrem*. A second frequent caesura falls after the first long of the fourth foot (hephthemimeral). The caesura gives the line its rhythm and often matches a break in sense; it does not change how you scan the line.

## Mistakes to watch for

- forgetting an elision;
- treating a short vowel before a single consonant as long;
- counting *h* as a consonant;
- splitting a diphthong into two syllables;
- forgetting that consonantal *i* between vowels counts double;
- not starting from the end, where the pattern is almost always the same.

Scansion also helps with meaning: in a scanned line *rosā* (ablative) can usually be told apart from *rosa* (nominative). That is one more reason to work through a poem yourself rather than read it only in translation, a point argued in [why translate it yourself](/en/blog/why-translate-it-yourself).

If you have a few lines of Virgil or Homer in front of you and are unsure of the forms, you can [try Lemma on your own passage](/en/demo): it works as [a Latin translator that explains every word](/en/latin-translator), and it does the same for [Ancient Greek](/en/ancient-greek-translator). Then take a pencil and mark the longs and shorts yourself, one foot at a time.

---

Source: https://lemma.algon.one/en/blog/how-to-scan-dactylic-hexameter · Lemma, the Ancient Greek and Latin tutor that cites its sources · https://lemma.algon.one/llms.txt