Up to hundreds of weaver ants…line up side by side in militarily precise rows. They grip the edge of one leaf with the claws and pads of their hindlegs and the edge of the other with their jaws and forelegs, and haul the two edges together. When the gap between the leaves is wider than the length of an ant, the workers use another, even more impressive tactic…: they chain their bodies together to form living bridges. The lead worker seizes a leaf edge with her mandibles and holds fast. The next worker then climbs down her body, grips her waist, and holds on. A third worker now climbs down to grip the second worker’s waist, and so on ant upon ant, until chains ten workers long or more are formed, often swinging free in the wind. When an ant at the end of the chain finally reaches the edge of the distant leaf, she fastens her mandibles onto it, closing the span of the living bridge, and all the entrained force begins to haul back in an attempt to bring the two leaves together. Sometimes the gap can be closed with a single chain, but usually several such large ensembles are needed, with nestmates working side by side. Some of the workers return from the site of activity to recruit nestmates by means of odor trails. They lay the trail substances not only over the leaves and twigs but over the bodies of the ants forming the chains. Soon a living sheet of ants is formed, and it presents a startling spectacle, its surface rippling with the slight movement of thousands of legs and antennae….

Now other weaver-ant workers move into position to apply the white “glue”…—threads of silk provided by the grublike larvae of the colony. How the silk is applied is the most amazing behavior of all in the repertory of the weaver ants, and the appropriate source of their vernacular name. The larvae recruited are in the final stages of development, following the last shedding of their skin as part of the growth process and prior to the next molt that will transform them into a pupa and inaugurate their changeover to the six-legged adult body form. In the nest-building process, such individuals are picked up by major workers, members of the larger of the two adult worker castes, and carried out to the leaf edges. Holding the larvae gently in their mandibles, the workers move their young charges back and forth across the leaf edges. The larvae respond by exuding threads of silk from a slit-shaped nozzle just below the mouth. Thousands of such threads stuck into place side by side spread as a whole into a sheet between the edges, in time to become a powerful adhesive that binds the leaves in place.

The choreography of silk spinning that ensues is a swift, precise pas de deux. The worker approaches the edge of the leaf while holding the larva in her mandibles so that the larva’s head projects well out in front, as though it were an extension of her own body. The tips of her antennae are brought down to converge on the leaf edge. For two-tenths of a second the tips play along the surface, not unlike the hands of a blindfolded person feathering the edge of a table to gain a sense of position and shape. Then the worker brings the larva’s head down to touch the surface. One second later she lifts it again. During this interval the worker vibrates the tips of her antennae around the larva’s head, touching it lightly about ten times. The subtle tapping is apparently a signal for the larva to release the silk. We are not certain that the movement contains such a command, but while it is occurring the larva does release a minute quantity of silk, which automatically sticks to the leaf surface.

An instant before the larva is lifted from the leaf’s edge, the worker raises and spreads her antennae. Then she turns her body and carries the larva directly to the edge of the opposing leaf, causing the silk to be drawn out as a thread. When she reaches this second surface, she repeats her earlier movements almost exactly. This time the larva touches the silk to the leaf and fastens the thread. Then both worker and larva return like tango dancers to the first edge to recommence the cycle. And so on metronomically, en masse, a rhythmic army of workers and larvae toils day after day, pulling together and sealing hundreds of pavilions across the great canopy empire. The ants add silken tunnels and rooms within the pavilions to create even tighter, more elaborate living quarters.

In 1964 Mary Leakey (doyenne of the Kenyan family of paleobiologists that has contributed greatly to our knowledge of the fossil history of man) sent Wilson a partial fossil colony of an extinct species of Oecophylla she had found during the search for early human remains. The age of the ant remains was approximately 15 million years. The fossils consisted of numerous fragments of different life stages and castes closely resembling those of the modern African and Asian weaver ants. The pupae were naked. That is, the larvae, like those of modern species, had spun no cocoons. Also, fragments of fossilized leaves were mingled with the ants. It thus appears that long ago a pavilion of weaver ants fell from a tree into a pool of water which was then covered by a rapidly congealing calcareous sediment. If that much is true, the unique social system by which Oecophylla weaver ants dominate the tropical canopy today appears to have been in place 10 million years before the origin of humanity.

—Bert Hölldobler and E. O. Wilson: Journey to the Ants (1994)

Photo sources here, here, and here.