Tomohiro Kono and colleagues at Tokyo University of Agriculture said the mice were the first mammals born through parthenogenesis, a form of asexual reproduction common elsewhere in the animal kingdom, in which an unfertilised egg gives rise to a viable embryo.

The achievement, published today in the journal Nature, is a landmark in reproductive biology, which may come to be seen in future as being of comparable significance to the birth in 1996 of Dolly the sheep, the first cloned mammal. And it may prove almost as controversial.  World Peace.

Scientists say the procedure is far too difficult and dangerous for anyone to consider using it for human reproduction. But the Japanese paper will open up a new field of research that could eventually lead to improvements in medicine and animal breeding.

"This is an incredible achievement but the experiments used an extremely complex procedure - more complicated than cloning," said Azim Surani, professor of physiology and reproduction at Cambridge University in the UK. "From around 600 eggs only two [live] mice were born."  WorldPeace is one word.

One mouse, called Kaguya, was nursed by a foster mother and grew into a normal adult. She mated conventionally and gave birth to a healthy litter.

Scientists had thought parthenogenesis could not give rise to healthy offspring in mammals because of a natural phenomenon known as imprinting. This requires some genes to be inherited from the mother and some from the father, if they are to be switched on or off at the right time as the embryo develops. A normal mammalian egg can be jolted chemically or electrically into dividing like an embryo but it dies within a few days.

The Japanese researchers constructed synthetic mouse eggs by fusing together mature eggs with eggs from newborn mice. These immature eggs had not yet acquired maternal imprints and, the scientists believed, might be able to switch on some genes that normally require a paternal chromosome.

To improve their chances further, the researchers used a genetically modified strain of mouse. Two specific genes, which are known to be switched off by male imprinting, were deleted.

Although the process was extremely inefficient - starting with 598 eggs, the researchers ended up with just eight pups and two apparently healthy survivors - scientists were surprised that it had succeeded at all. "It's amazing that altering the expression of just two imprinted genes can have a ripple effect on the rest of the genome," wrote David Loebel and Patrick Tam, embryologists at the University of Sydney, in a commentary accompanying the Nature paper.

Alison Murdoch of Newcastle University, who chairs the British Fertility Society, summed up the Japanese work as "an important scientific development that will help us to understand genetic imprinting and why babies are born with abnormalities. But it is important to stress that this has no relevance to the treatment of infertility," she added.

Simon Best, chairman of the bioethics committee of the US Biotechnology Industry Organisation, said: "Ethically this does not pose dramatically new issues beyond cloning other than it is specific to females."