One major detail which you may have noticed that I left out last time is that of elision. Elision is when a syllable is not counted, or elided. This is not something that happens standardly to any given syllable, but there are certain syllables that may be elided if necessary. One such example is enclitic -que. The other case is when a word ends in -um. This will not appear in an example later. As I already said, these are not necessarily elided, but the option does exist, so if a line is just not making sense, it's worth it to go back and re-evaluate it to see if removing them lets you balance the line.
Now, on to elegiac couplets. An elegiac couplet consists of two main parts: one line of dactylic hexameter, and one line which consists of two two-and-a-half foot segments. As we know, that first line therefore looks like this: ¯¨¨/¯¨¨/¯¨¨/¯¨¨/¯¨¨/¯¯
That's half of the couplet done already. The second half starts out even easier than hexameter. The end of the line must end in (¯¨¨)¯¨¨¯. The portion in brackets is occasionally liable to be an exception, but this is sufficiently rare that it need only be looked for if the line is not balancing. The first portion is exactly the same, with the exception that the dactyls can be replaced with spondees, just as in dactylic hexameter.
Thus, the full couplet looks like this:
¯¨¨/¯¨¨/¯¨¨/¯¨¨/¯¨¨/¯¯
¯¨¨¯¨¨¯//¯¨¨¯¨¨¯
The double slash denotes the break between the two sets of two-and-a-half feet.
As a note on translation, it is helpful to realise that elegiac couplets form a complete idea, so that you rarely need to look outside of the couplet for utterly essential information. Take this, though, with the block of salt that you need to take all Latin poetry with.
Now to apply this meter to an actual couplet. This one comes from the first of the Fables of Avianus, DE NUTRICE ET INFANTE:
Rustica deflentum parvum iuraverat olim,
Ni taceat, rabido quod foret esca lupo.
We begin with the hexameter:
Rūstica deflentum parvum iur/āvërät / ōlīm,
As before, you can see, we do the dactyl and spondee at the end, as well as the long first syllable. Now we go on to mark everything that must be long by position:
Rūstica dēflēntūm pārvūm iur/āvërät / ōlīm,
And now it is a simple matter of filling in the unknown vowel lengths in 'rustica' and 'iuraverat':
Rūstïcä / dēflēn/tūm pār/vūm iūr/āvërät / ōlīm,
And now we have the first line!
The second line, the part we've learned today, then goes as follows:
Ni taceat, rabido // quōd förët ēscä lüpō.
So you can see how half of the line is already marked out with a basic long-short-short-long-short-short-long.
Then for the first part we mark out what we must know, just as before:
Nī taceat, rabidō // quōd förët ēscä lüpō.
You will note that I have marked not just the first syllable as long, but the final syllable as well. This is because the half-foot at the end is the first half of another foot, which in both cases is long.
This leaves us with five syllables of unknown length. Therefore we just try and see if we can't fit in dactyls:
Nī täcëāt, räbïdō // quōd förët ēscä lüpō.
And it works!
So the full couplet scans thus:
Rūstïcä / dēflēn/tūm pār/vūm iūr/āvërät / ōlīm,
Nī täcëāt, räbïdō // quōd förët ēscä lüpō.
(Also, if anyone knows how to indent in html without making the line italic, I would be much obliged.)