Restored

The returned member of the parent’s folder — a name born in 1977, sunk under another species without a published reason, and brought back in 2026 without its types in hand

R990 closed by pointing here: Temnothorax caucasicus (Arnoldi, 1977), sunk for years under Temnothorax nadigi (Kutter, 1925) and restored as a good species in the same 2026 revision of the bulgaricus species-group from the Aegean Region (Annales Zoologici 76) that described mersinensis new. The question was what a restored name holds that a new name does not. All of what follows is summarised in my own words, with no quoted text beyond bare collecting data.

First, this name had a first life, and it was a real one. In my summary of the original record: Arnoldi described it in 1977, in Leptothorax, from Georgian material from Kojori and Tsagveri — a holotype and three paratypes — together with four workers from Jrvezh in Armenia; the revision gives Tsagveri, Georgia, as the type locality. It was filed at birth in the clypeatus group, on the strength of two features it still carries: a seam across the promesonotum and a notch in the middle of the clypeus. In general build it was said to resemble clypeatus, set apart by a low, rounded petiole. A new name, like mersinensis, holds only its description. A restored name holds a description plus everything that was printed under its own spelling before anyone took the spelling away.

Second, the taking-away was never argued in print. In 2002 the name was listed as a junior synonym of nadigi, credited to a colleague’s work in preparation. No published justification for that sinking ever followed — the 2026 discussion says so in substance, and the comments under the species say the same. So for roughly two dozen years the ant lived in the literature inside another species’ record, not because a case against it had been made and won, but because a case had been announced and never delivered. That matters for what restoration means here: the 2026 authors did not overturn a verdict. They closed a file that had been left open, as a note, since 2002.

Third, the bringing-back was done without the types. The revision states plainly, in my summary, that direct examination of the type material was not possible because the types were inaccessible. The restoration rests instead on Arnoldi’s original description and its documentation, judged clear enough to recognise the animal by, and on newly examined material that matches that diagnosis without overlapping its relatives. The photographed workers, queen and male in the account come from a village in Yozgat Province, Türkiye — the living end of the evidence, standing in for pins the authors could not reach. A restored name, then, can hold a debt a new name never carries: its anchor specimens are cited, not seen, and the match between the 1977 words and the modern material is itself part of the claim.

Fourth, what separates it from the name it was sunk under is sculpture, size and the queen — not a clean measurement gap. Together, caucasicus and nadigi stand apart from the group’s other three members: bodies in two tones with the head visibly darker than the mesosoma, a sculptured back of the head, and both the clypeal notch and the promesonotal seam present. Against nadigi itself, caucasicus is told, in my summary, by a head built of distinct ridged wrinkles with finely netted spaces between them, where nadigi is mostly dense fine netting with only faint ridges; by stronger, thicker mesosomal sculpture whose spaces are usually smooth, where nadigi’s are netted; by plainly larger workers on the revision’s mean measures; and by queens that are predominantly yellow with the side plates of the mesosoma sculptured across their whole surface, where nadigi’s queens are mostly reddish to dark brown with those plates smooth in front. And yet the numbers cross: in the revision’s discriminant test, which sorted the group’s workers correctly about nineteen times in twenty overall, three of the seven caucasicus workers in the test were sorted elsewhere — two of them as nadigi itself, one as cypridis. The page keeps that overlap in full view. The restoration is a morphological judgement the measurements support only in part, and the paper itself notes that this trio of species differs clearly in build, range and queens despite what the indices fail to separate.

Fifth, the animal underneath is an upland one, and its map is the eastern edge of the folder. In my summary of the biology, drawn in the account from earlier literature: an alpine species of riversides and deciduous forests between about 1,000 and 1,900 metres, nesting in branches, under bark, in soil and under stones. Its listed distribution is Georgia, Armenia, Azerbaijan and Türkiye — the key narrows the Turkish part to the east. Compare its neighbours in this sequence. Mersinensis holds one point and no past. Bulgaricus holds the indexing twice. Caucasicus holds a range of four countries plus a past life in print — and, uniquely here, a share of someone else’s map: the revision’s nadigi account lists that wide species from the Caucasus region, Iran and Turkmenistan, and then flags exactly those eastern records as needing reinvestigation, because they may belong to caucasicus instead. Restoration does not just give a name its own records back. It puts a question mark inside the record of the name that hosted it.

So what does a restored name hold that a new name does not? A new name holds a promise. A restored name holds a promise, a prior life, and an unpicking job: every record filed under nadigi from the Caucasus eastward now has to be asked, one by one, which of the two animals it actually was. It also holds two honest weaknesses, kept at their full height on this page: it was sunk without a published reason, and it was raised without its types being seen. Both acts ran on less than the ideal evidence, in opposite directions. What tips the second is that its reasons are printed, checkable, and hedged in the open. That reading is this page’s own framing, not a claim in the revision.

A schematic walk — five stages of a name that returned. Schematic only: it draws the shape of the record, not the geography. Click a stage.

1 · born under its own spelling — 1977
2 · sunk on a note, never argued — 2002
3 · filed inside another name — 2002–2026
4 · restored, types unseen — 2026
5 · four countries, one unpicking — still

Click a stage.

The newest member waited in a drawer for a folder to reopen. The restored member waited inside a folder that was never really closed — a synonymy announced, never argued, and finally withdrawn by authors who could not even reach its types. The drawer kept mersinensis safe. Only the unpicking can say how much of nadigi’s east was caucasicus all along.