In 1699 Maria Sibylla Merian sold studio goods in Amsterdam, boarded a merchant ship with her daughter Dorothea, and spent two years in Suriname doing something European cabinets rarely asked of a moth: keeping it alive long enough to watch it change.
She was fifty-two. She took jars, brushes, and the patience to feed caterpillars their actual plants. Spontaneous generation (the idea that insects cooked themselves out of slime and rotting meat) was still a respectable enough error that her daily notes were an argument, not a hobby. When Metamorphosis insectorum Surinamensium appeared in 1705, the shock was not only tropical color. It was method. Merian did not pin a specimen in a blank void. She drew the insect on the plant it ate, and sometimes on the animal that ate it.
The famous plate of a bird-eating spider with a hummingbird has been argued over ever since: observed behavior, composite, or a claim too strong for the evidence. That argument is useful. It reminds you that a natural-history plate is a made object, not a photograph, and that the making was already a fight about what counted as seeing.
Copper, Reverse, and a Second Set of Hands
Before photographic reproduction, the published insect was a metal problem.
An engraver cut the drawing into copper with a burin, in reverse, because the press would flip it. Latin names, scale bars, and the vein of a leaf all had to be thought backward. A right-handed moth on the watercolor became a left-handed moth on the metal unless someone remembered to flip the design first. Ink was pushed into those cuts and printed onto sized rag paper. Color, when the buyer paid for it, was often a later human pass: mineral and insect pigments laid into the hatching by hand. Ultramarine, cochineal red, and ochres did not come from a software swatch. They came from mines and boiled insects, and they sat in the engraved lines like weather. Two impressions of the same plate can disagree about a leaf because two colorists disagreed after dinner.
The plate could hold a line finer than casual looking. That is why a dragonfly eye or a moth thorax in a good impression still feels more specific than a lot of digital creature design. The information was expensive. Nobody spent a week cutting a wing to be vague.
The published moth is a relay. Someone watched a living insect and made a watercolor. Someone else, or the same person on a different day, translated that sheet into reversed metal. A third pair of hands, often unnamed, put color into grooves they had not cut, of a creature they had not fed. Each stage can correct or betray the last. That is how a plate becomes an argument: not because it is pretty, but because several people have decided, in sequence, what the viewer is allowed to know.
Merian’s Suriname work sits in a longer northern European practice of women and men who painted insects and plants as working knowledge, not as decoration for a margin. What she added, fiercely, was life cycle plus ecology on one sheet: hole in the leaf, frass, chrysalis, adult, predator. The page is a small weather system.
Haeckel, Glass Skeletons, and a Careful Afterlife
Two centuries later Ernst Haeckel’s Kunstformen der Natur (1899–1904) made the microscopic look like architecture. Working from marine samples, including material associated with deep-sea collecting of the Challenger era, he drew radiolarians: single-celled organisms whose silica skeletons read as rose windows, gates, and clocks. The plates are lithographic performances of symmetry. They are also Haeckel: a scientist-artist with a system, not a neutral camera.
Later decorative art used those plates. René Binet’s monumental gate for the 1900 Paris Exposition is the usual cited case of a designer taking Haeckel’s geometries into iron. That is a documented design conversation. It is not proof that Art Nouveau “came from” radiolaria, and it is not a reason to treat Haeckel as a saint of seeing. Some of his evolutionary pictures elsewhere have been criticized as too willing to make the specimen fit the theory. The Kunstformen plates remain extraordinary objects. They should be used as plates, not as a creation myth for a whole style.
Paleoart has a different job: reconstructing what is already bone. Natural-history illustration of living or newly collected organisms has this job: decide what to include in the same frame so the viewer understands a relationship. Merian’s wasp and caterpillar, or a frog in a pitcher plant, are essays in dependence. An invented monster that ignores food, molt, and habitat is usually a costume.
Museum libraries have put Merian and Haeckel online, which is a kindness and a flattening. The fantastical part was never the ink. It was that someone thought a real leaf, and the hole chewed in it, was worth a week of copper.