Our Milky Way smashed into its neighbouring Andromeda galaxy around 10 billion years ago, European astronomers suggest.
Previous studies have suggested that our galaxy is set to crash into Andromeda in 3-4 billion years, and that this will be the first time such a collision has taken place.
However, now a European team of astronomers led by Hongsheng Zhao of the University of St Andrews propose that the two star systems collided some 10 billion years ago and that our understanding of gravity is fundamentally wrong.
This would neatly explain the observed structure of the two galaxies and their satellites, something that has been difficult to account for until now, researchers said.
The Milky Way, made up of about 200 billion stars, is part of a group of galaxies called the Local Group. Astrophysicists often theorise that most of the mass of the Local Group is invisible, made of so-called dark matter.
Zhao and his team argue that at present the only way to successfully predict the total gravitational pull of any galaxy or small galaxy group, before measuring the motion of stars and gas in it, is to make use of a model first proposed by Professor Mordehai Milgrom of the Weizmann Institute in Israel in 1983.
This modified gravity theory (Modified Newtonian Dynamics or MOND) describes how gravity behaves differently on the largest scales, diverging from the predictions made by Newton and Einstein, researchers said.
Zhao and his colleagues have for the first time used this theory to calculate the motion of Local Group galaxies. Their work suggests that the Milky Way and Andromeda galaxies had a close encounter about 10 billion years ago.
If gravity conforms to the conventional model on the largest scales then taking into account the supposed additional pull of dark matter, the two galaxies would have merged.
"Dark matter would work like honey: in a close encounter, the Milky Way and Andromeda would get stuck together, figuratively speaking," said team member Pavel Kroupa from Bonn University.
"But if Milgrom's theory is right, then there are no dark particles and the two large galaxies could have simply passed each other thereby drawing matter from each other into long thin tidal arms," said Dr Benoit Famaey from Observatoire Astronomique de Strasbourg.
"New little galaxies would then form in these arms, a process often observed in the present-day universe," added team member Fabian Lueghausen, also from Bonn.
"The only way to explain how the two galaxies could come close to each other without merging is if dark matter isn't there. Observational evidence for a past close encounter would then strongly support the Milgromian theory of gravity," Zhao said.