Silent twice

The second parent turns out to be silent too — but its silence has a different shape: this parent was redescribed in full, measured fifty-three workers deep, and still never once names the subordinate it held for sixty-three years

R996 closed by pointing here: Temnothorax fumosus spent most of its filed life, 1927 to 1990, as a subspecies of Temnothorax bulgaricus, and the question was whether this second parent’s record mentions it any more than the first parent’s did. R989 already gave bulgaricus its own page in this sequence; this page answers the narrower question that page left open, and it does not revise it. Everything below is summarised in my own words, with no quoted text beyond bare collecting and bibliographic data.

First, what the parent is, freshly and fully described. Forel described it in 1892, at page 314 of the first paper on Bulgaria’s ants, in the old genus Leptothorax, from Sliven. The 2026 revision of its species-group redescribed it completely: its lectotype is a Sliven worker in Geneva (CASENT0909014), labelled as such in 1999 and personally examined for the revision, with a paralectotype in Genoa (CASENT0904743); fifty-three workers and ten queens were measured. In the revision’s statistical test, the misclassified workers all fell among three of the other members — bulgaricus itself was never among the specimens the analysis put in the wrong species.

Second, what it was keyed against. The 2026 key first places it on the pale, seam-free branch of its group: head and body mostly one yellow-orange tone, no promesonotal suture, no notch in the clypeus, head sculpture fading toward the back — the branch it shares with cypridis and mersinensis, against the darker, seamed, notched pair caucasicus and nadigi. Within that branch it is separated from the Cypriot by its much longer erect hairs (pronotal setae usually over a tenth of a millimetre, against the islander’s shorter coat) and from the Mersin species by a shorter antennal scape relative to the head and by a different pattern of grooves on the head. Its range in the revision is the group’s widest: Bulgaria, Greece, North Macedonia, Serbia and western to central Türkiye, with an Israeli record the authors flag as needing confirmation; it lives mostly in oak, mixed or coniferous forest, and nests in rock crevices and dry twigs.

Third, the swarm it once held. Across its life this one name carried, at one time or another, at least five subordinates that are all species in their own right now: the three 1911 names graecus, aeolius and smyrnensis, born underneath it and later promoted into a group of their own; cypridis in 1930, promoted inside bulgaricus’s own group a lifetime later; and fumosus, moved under it in 1927 and raised out of it in 1990. Every one of those promotions is recorded somewhere — in the subordinate’s own record.

Fourth, the answer. I searched the full text of the 2026 revision — the one work that redescribed this species, restored one member of its group, described another new, anchored a third with a lectotype and put a fourth out of the group altogether. The name fumosus does not occur in it. Not in the species account, not in the synopsis, not in the key, not anywhere. The second parent’s fullest modern record is exactly as silent about the subordinate it held for sixty-three years as the first parent’s catalogue entry was about the four years it held it.

But the two silences are not the same silence. Satunini was never redescribed at all, so its silence could be mere neglect. Bulgaricus was redescribed with its types in hand — and fumosus is absent for a structural reason the revision states openly: the work is a revision of a group, inside a drawn map, the Aegean region of Greece, Cyprus, western and central Türkiye, Lebanon and western Syria. Fumosus was born on Cheleken Island, in Turkmenistan, far east of that map, and it belongs to no Aegean group the revision treats. A former child that has left both the folder and the map does not get mentioned when the folder is cleaned — not out of disregard, but because a revision’s table of contents is a border, and borders are exactly what a promoted name crosses.

The walk — five moments in one parent’s record, schematic only. Click each stage.

born at Sliven, 1892
parent of five, 1911-1930
last subordinate leaves, 1990
redescribed in full, 2026
unnamed in its own revision, 2026

Pick a stage.

So what does the second parent hold? Everything about itself, in unusual completeness — an examined anchor, a clean statistical outline, the widest map in its folder. What it does not hold is any trace of the names that once lived under it. Parenthood in this filing system is apparently a one-way record: the child’s page points up forever; the parent’s page, even at its most thorough, points only at itself.