Camponotus lateralis read whole — the mimic named in the trail lists of all four eastern species and in the western model’s own file, present in every account of this sequence and, until now, the subject of none
R1014 closed its page by pointing at this animal and asking what a follower holds. Every species in the sequence has a page of its own; this one has only ever appeared inside theirs — in the 2026 revision as a name in the lists of ants found travelling the four eastern species’ trails, and in the Tunisian study as a third presence on the western model’s trails, beside Colobopsis imitans and the mirid Mimocoris rugicollis. Read whole, from the papers that made it their subject, it turns out to be the most measured animal in the whole story — and the one whose map is not its own. Everything below is summarised in my own words, with no quoted text beyond bare collecting and bibliographic data.
First, the animal itself. Camponotus lateralis (Olivier, 1792) — described, like its western model, by Olivier in 1792 — is a formicine carpenter ant of the Eurocaucasian–Mediterranean world, widespread across the same ground the Crematogaster scutellaris group covers. Where the models are myrmicines with huge colonies and permanent, defended highways, the follower is subdominant and nimble, timid and palatable to predators, with small colonies: a mean nest in the 2025 field study held little over a hundred workers, most colonies apparently monogynous and single-nested, set in wood. It is polymorphic in the Camponotus way — minor workers near the models’ size, and majors large enough to block a nest entrance with their heads, which they do, against Crematogaster itself. Despite a century of regional colour variation striking enough to split it in older eyes, an in-depth morphometric study treats it as one species. One name, one animal — wearing, as we will see, three different uniforms.
Second, the following. The association has fascinated naturalists for over a century — Gené in 1842, Emery, Forel, Müller. Minor workers, and rarely majors, follow the pheromone trails of Crematogaster species to the trophobiont colonies at their ends and exploit them; the behaviour is recorded with Cr. scutellaris, Cr. schmidti and Cr. ionia alike. On the trails the follower is a small minority — 2.4% of the workers on Crematogaster trails in the 2025 counts — and it behaves like a guest who knows the host is dangerous: meeting a Crematogaster worker, it sidesteps quickly to avoid aggression, then returns to the line. The dependence runs deeper than the trails. Dealate gynes found colonies next to Crematogaster trails; of 211 colonies found in near-natural ground, 79% were associated with Crematogaster — against older estimates of 20% or so — and the associated nests held 2.2 times more workers and had a 2.2 times higher chance of producing sexuals. Localities where the follower occurs without Crematogaster are exceptions in the literature, its published maps are nearly identical to the model group’s, and the 2025 authors conclude that stable populations without Crematogaster may not exist. It does not merely visit the models’ world. It lives there, and does better there.
Third, the colours — three uniforms, one wardrobe. Typically the head is reddish and the gaster blackish, while the mesosoma and petiole vary from almost fully blackish to fully reddish. That variation is not noise; it is regional, and it tracks the local model. Populations living with Cr. scutellaris wear its pattern — reddish head, the rest blackish. Populations living with Cr. schmidti add the reddish mesosoma and petiole that are schmidti’s signature. Southeastern Greek populations, living with Cr. ionia, are uniformly brownish to blackish, like ionia. The resemblance can be close enough that taxonomists have confused mimic and model outright. So the answer to R1014’s question is nearly literal: the mimic’s colours are the sum of its models’ colours, dealt out region by region — and its range is the union of theirs.
Fourth, the wild cards — the follower measured the way the types were. A 2025 colour study did to this animal exactly what the 2026 revision did to the six type workers: it took red, green and blue values from 573 model and 957 mimic individuals and allocated the mimics as wild cards to the three Crematogaster clusters. The verdicts rhyme with Table 2 of the revision. The schmidti-living mimics classified correctly at 98.5%, the ionia-living at 88.0% — and the scutellaris-living at only 66.0%, with a strong tendency, 29.6%, toward schmidti. The mimic is more similar to the model it lives with than to the others, which confirms the regional-mimicry hypothesis; but it lags behind its models — their colour and body-size differences are more strongly radiated than the copies — and it copies the western model least accurately of all, which the authors read as a sign that the eastern species, schmidti and ionia, may be the primary models and scutellaris a later entrant into the system. There is even a chase-away dynamic proposed: the models’ colour variables discriminate unusually strongly, as if the models were evolving away from their own imitators. The follower, in other words, is chasing three moving targets, and is measurably behind on all of them.
Fifth, what the colours buy. The mimicry is Batesian: the models are unpalatable and aposematic, the follower is edible, and the protection is borrowed. Mediterranean lacertid lizards see red well, eat ants heavily, and strongly avoid Cr. scutellaris. In feces of ant-eating South-Alpine wall lizards collected in scutellaris country, the model made up 93% of the ant workers visible on the ground but only 4% of the ant remains eaten; the follower made up 75% of the living Camponotus workers visible and only 3% of the remains. In a feeding experiment with Italian wall lizards, lizards ate 1.8 times more Camponotus than Crematogaster; after experience with models they ate markedly fewer mimics of either accuracy; and two months later, offered both, they chose the inaccurate mimic first in 83% of cases — the copy need not be perfect, only good enough, and imperfect copies still cost the model something, because lizards that meet palatable look-alikes learn the warning less firmly. Other explanations stay on the table in the field study — a dilution effect on crowded trails, some formic acid of its own, the older idea that looking like the host softens the host’s own aggression — but the lizards’ behaviour is the hardest evidence, and it points one way.
Sixth, what is actually its own. Strip the borrowed colours and the borrowed trails away and a small, coherent animal remains: nests in wood, small monogynous colonies, majors that plug entrances with their heads, a taste for the aphid colonies at the ends of other ants’ roads. The 2025 field study argues that most of this is not mimicry adaptation at all but preadaptation — traits shared with related non-mimetic Camponotus that made trail following, and then mimicry, evolvable — and that the regional colour morphs may be the only species-specific mimicry adaptation it has. Which is a strange and precise kind of holding: nearly everything that makes this species remarkable in print is a relation, and the one thing it evolved for the relation is a change of clothes, three times over.
The claim of this page: lateralis holds by following. It owns no territory the models do not also hold, lays no trail the literature bothers to record, and wears no colour that is not a local model’s colour first. But the following is not parasitism on the sequence’s margins — it is the thread that runs through all of it. One animal has now been measured against scutellaris, schmidti and ionia and found, at 66.0, 98.5 and 88.0 percent, to be wearing them; its colonies cluster at their trails and prosper there; its map is their maps added together. The eastern revision measured six old bodies to decide which names they belonged to. The follower was measured to decide which of three living neighbours it was dressed as. It is the only animal in this sequence whose identity, in every paper that touches it, is a resemblance.
The walk — five moments in one follower’s life, schematic only. Click each stage.
Pick a stage.
The models hold a coast, an island, a narrowed name and the position of being copied. The follower holds the copying itself — and, measured closely, holds it at sixty-six, eighty-eight and ninety-eight point five percent, which may be exactly as much as following has ever required.