Denmark's coasts are dotted with long, low mounds of shell — oyster mostly, mixed with mussel and snail, and studded here and there with bone and worked flint. For most of the nineteenth century, respectable opinion held that these were natural features: shell banks thrown up by storms and currents, then stranded as the sea withdrew. In 1848 a commission was formed to settle the question, and its composition was ahead of its time. The geologist Johan Georg Forchhammer (1794–1865) would read the deposits. The zoologist Japetus Steenstrup would read the shells and the animal bones. The archaeologist Jens Jacob Asmussen Worsaae (1821–1885), who had already used stratigraphy to prove Thomsen's Stone–Bronze–Iron sequence, would read the artefacts. In 1851 the three of them dug at Meilgaard.

The mounds were rubbish. Nobody had piled these shells; people had eaten the animals inside them and thrown the shells aside, meal after meal, generation after generation, until the discard became a landform. The Meilgaard heap belongs to the Ertebølle culture, the last hunter-fishers of the Danish Mesolithic, who worked these shores roughly between 5400 and 3950 BC. What the commission demonstrated, though, was bigger than one conclusion about one heap. It was a method: bring the geologist, the zoologist and the archaeologist to the same trench and let each read what only they can. Steenstrup's word for the mounds crossed into English, and stayed.

Danish archaeology came back to the question forty years later with far more rigour. Between 1893 and 1895 a second commission, run by the National Museum, excavated 314 square metres at Ertebølle itself on the Limfjord and recovered 8,608 artefacts, 20,300 animal bones and 563 pieces of charcoal — the kind of counting that turns an argument into data. The results appeared in 1900 as Affaldsdynger fra Stenalderen i Danmark, 'Refuse Heaps of the Stone Age in Denmark'. The Meilgaard midden, meanwhile, still lies where it always has, in the coastal forest of Nederskov just north of the castle, quiet under the trees and largely unvisited.

Meilgaard the estate is far younger than the science practised in its woods, though old enough by most standards. It appears in the record as Medelgård in 1345, when it changed hands. The three-winged main building went up in 1573, was enlarged in 1873, and was then thoroughly rebuilt between 1888 and 1891, when the architect August Klein added a storey. Its list of owners runs across five and a half centuries of Danish surnames — Skjernov, Juul, Kaas, Bille, Rosenørn, Sehested — a reminder that a manor is less a building than a relay race in which the baton happens to be a house.

On 26 February 2003 Meilgaard burned. The main building was reconstructed between 2004 and 2006, which is why a house first raised in 1573 now has interiors younger than most of the people who visit it. The Iuel family has held the estate since 1868, when Baron Christian Frederik Theophilus Alexander Juel-Brockdorff (1827–1888) bought it; it passed to Michael Frederik Iuel in 1986, and he and his wife, Princess Désirée of Schaumburg-Lippe, own it still. In the wood just north of the house, the shells have not moved since the Stone Age. They have now outlasted at least one version of the house that made them famous.