Two-hundred-year-old seeds found at the National Archives have sprung back to life following careful coaxing from experts at Kew Gardens.

The ancient seeds have unexpectedly started to grow into vigorous young plants after they were germinated by seed scientists from the Millennium Seed Bank at Wakehurst Place, the Royal Botanic Gardens, Kew's garden in West Sussex.

Although botanists at Kew Gardens are used to planting seeds and letting them grow, the team has never been asked to use such old seeds.

The 32 different species of seeds in 40 small packets were discovered by Roelof van Gelder, a guest researcher from the Royal Dutch Library, in a red leather-bound notebook within files held at the National Archives.

The notebook was inscribed with the name Jan Teerlink, a Dutch merchant who is believed to have collected the seeds during a trip to the Cape of Good Hope in 1803.

Germination was going to be tricky but not impossible once colleagues from the Millennium Seed Bank were called in to help. Now three of the 32 species have germinated and the ancient specimens are growing into healthy plants in the glasshouses at Wakehurst Place.

Seed ecologist Matt Daws said: "This is a fantastic result. The seed was so old and had been stored in some dubious conditions, including a ship and the Tower of London. We really did not expect to get anything."

Alistair Hanson, an early modern specialist at the National Archives, said: "This is an exciting discovery and a testament to the hard work that goes into preserving the files at the National Archives.

"To be able not only to discover these seeds but also to germinate them helps to bring history to life, literally. I will be keeping my eye on the seeds' developments over the coming months and years."

The first seeds to germinate belonged to the legume Liparia villosa.

Of the 25 seeds Daws planted, 16 sprouted. The second was labelled Protea conocarpa on the original packet, although Kew's scientists have now identified it as almost certainly a species of Leucospermum, which is of the Proteaceae family. Just one out of eight seeds of this species germinated.

The exact identity of the last of the trio remains a mystery, although the team know it to be a second legume, this time an Acacia.

"We'll have to wait until it flowers to find out what species it is," said Matthew. "If it's a tree, we may have a long wait."

This sample consisted of just two seeds. One germinated and is now half a metre tall.

The second failed and microscopic examination revealed old insect damage.

For Kew's scientists this project has been of more than historical interest.

"According to models of seed survival, even the toughest cereal seeds should have died after so long in such a condition," explained Matt.

"If seed can survive that long in poor conditions, then that's good news for those in the Millennium Seed Bank stored under ideal conditions."

The seeds were carbon dated by Kew's science team to verify their age and Matt's colleagues are now extracting DNA from live and dead seeds to complete the study.