Sapphic metre in Catullus 51.10

Question

I am trying to understand the metre of Catullus 51, and this line has me baffled:

tintinant aurēs, gemina teguntur

The standard sapphic eleven-foot metre is – ⏑ – ⏓ – ⏑ ⏑ – ⏑ – –, that is 𝅘𝅥 𝅘𝅥𝅮𝅘𝅥 (𝅘𝅥/𝅘𝅥𝅮) | 𝅘𝅥 𝅘𝅥𝅮𝅘𝅥𝅮| 𝅘𝅥 𝅘𝅥𝅮𝅘𝅥𝅘𝅥. This should therefore map as follows:

tintinant aurēs, gemina teguntur
 –  ⏑ –    – –    ⏑ ⏑ –   ⏑ – –

As I understand gemina, it is either referencing Catullus’ eyes (previously unmentioned), but I would then have expected [oculī] *geminī teguntur; or they are pointing back to the ears, which does make sense grammatically (aurēs gemina¹ teguntur). I cannot see how gemina can be ablative singular. What is going on in this line? Why is gemina (apparently) three breves? Is it possible in sapphic metre to have ⏑ ⏑ ⏑ in that position?

───

1 Rookie mistake; were it that, it should obviously have been aurēs geminæ as per the accepted answer.

Answer 1

As a first note, the third and fourth lines of a Sapphic stanza tend to be closely connected—it's not uncommon to have a word split between them—so let's add that fourth line in here.

tintinant aurēs, gemina teguntur / lumina nocte

Now we can see why it's gemina rather than geminī: it's modifying not an implied oculī, but an explicit lumina.

The scansion, though, doesn't quite line up. As you say, that syllable should be heavy. I've seen this analyzed in a few different ways.

First, it could be geminā, with a long ā. In this case it would agree with nocte rather than lumina; "twin night" doesn't make much sense, but it can be analyzed as a transferred epithet. Adjectives agreeing with the wrong nouns isn't uncommon in poetry—though this would be the only instance of it in all of Catullus—or it could be some metaphor, he's covered by both literal and metaphorical nocte. The Loeb edition takes this route.

Second, it could be geminae, with a diphthong. In this case it would agree with aurēs rather than lumina; "twin ears" makes more literal sense than "twin night", at least, but this feels a bit less elegant to me, since there's a natural caesura between aurēs and gemina. Schrader apparently prefers this version.

Third, it could be gemina et "and my twin eyes". This lets it agree with the most obvious noun, lumina. I've seen this option taken in some online editions, though the Loeb doesn't mention it. I'm going to dig out my physical edition later today and see if the apparatus has anything to say on the matter.