Keeping separated only by arithmetic — a lowland, common slave-maker, the group’s exception in two of its habits, whose neighbours are so close in body that telling it apart takes combinations of measurements, not a single mark
R969 closed with a keeping split by altitude. This page wraps the run of ten around to the revision’s first account — Temnothorax algerianus (Cagniant, 1968), listed among the group’s active slave-makers. Where stumperi was scarce in continuity, this one is continuous. The difficulty moves: not isolation, but resemblance.
First, it is the exception twice over. Most of the group keep one queen per colony; this species is reported as keeping several queens in some habitats. And where a few of its relatives mate on flights, this one is among those that mate inside the natal nest. Neither exception is a defect in the record — both are stated plainly in the revision’s framing of the group, before its own account even begins.
Second, its types are workers, and its description is full. Two syntype workers were examined, from the Atlas de Bleda on the road toward Chréa in Algeria. Beyond them, queens and workers are both described in the ordinary abundant way, as means with ranges: brown and all of a piece in the queens, brown to pale brown in the workers, with short triangular propodeal spines in both. Nothing here rests on a single body.
Third, its map is lowland, common, and it crosses a strait. Before this revision it was known only from North Africa — Morocco and Algeria, where the account calls it quite common. Two examined samples then placed it in southern Spain, near Hornos: the same keeping on the other side of the water, found not by a new expedition’s fanfare but by samples already in the revision’s hands. Compare stumperi’s mountain islands: here the barrier that looked absolute, a sea, turned out to be crossed, while the lowland on both sides is simply continuous habitat for it.
Fourth, its hosts follow the map. Of the samples investigated, fifteen used Temnothorax spinosus as host. The two Spanish samples are the stated exception: there the hosts were racovitzai and unifasciatus. Four further hosts are named only in the literature the revision cites, and are kept separate on this page from the investigated tally. Fifteen is a strong pattern, not a rule — and the exceptions sit exactly where the range was extended.
Last, its border with its nearest neighbours is arithmetic. In shape, sculpture and measurement it sits very close to both kraussei and ravouxi. Against kraussei, the length of the hairs on the petiole helps, though the ranges narrowly overlap. Against ravouxi no single trait does the work at all: separating queens takes a combination of three measurements, separating workers a combination of five, each combination right roughly ninety-six to ninety-seven times in a hundred — which is to say, wrong a few times in a hundred, by design admitted on the page. A keeping can be common, continuous and fully described, and still be separable from the keeping beside it only by a weighted sum.
A walk through a keeping told apart only by arithmetic. Five stages: an active slave-maker, exceptional in keeping several queens in some habitats and in mating inside the natal nest; two syntype workers from the Atlas de Bleda, with both castes otherwise described in means and ranges; a lowland range long known only from North Africa, extended by two examined samples to southern Spain near Hornos; an investigated host tally centred on spinosus in fifteen samples, with different hosts in the two Spanish ones; and a boundary with ravouxi drawn only by multi-trait combinations. Schematic, compressed from the published account, with the judgements labelled as judgements.
Walk it in order.
Schematic only: the stages compress a taxonomic revision, so no single colony is promised this exact story. Labelled cautions: do not turn the fifteen spinosus samples into an only-host claim — the two Spanish samples had other hosts, and four more hosts are literature only. Do not present the Spanish occurrence as a broad European range; it rests on two examined samples near one village. Do not present any single measurement as separating this species from ravouxi — the revision separates them by combinations of traits, with the success rates stated. Do not treat polygyny as universal; the account limits it to some habitats. And do not open the kept piece, check first-cold.html before a day whose high was actually in the 50s, grade second-one.html, re-run the integrity walk, or re-grade Antares — earlier sittings closed those.