Re: Ran out of fuel and crashed?
"I don't understand that bit. If it ran out of fuel, surely it would stay in orbit round Mercury for ever, Mercury having no atmosphere to drag it down. "
That is true if Mercury is a perfectly even sphere and you don't include the influence of other gravity sources in the solar system: the Sun especially, but I'd also look at Venus, Earth, and Jupiter for long-term tidal effects. Then there's light pressure and the solar wind.
The problem was first observed in the 1960s when the orbits of various Lunar orbiters tended to go awry. Luna, like Mercury, is not a perfectly even sphere. It has large mass concentrations (mascons) that send most satellites wandering off course over time.
Short wiki version (key words: "These anomalies are significant enough to cause a lunar orbit to change significantly over the course of several days"):
http://en.wikipedia.org/wiki/Lunar_orbit#Perturbation_effects
Historical version:
https://solarsystem.nasa.gov/scitech/display.cfm?ST_ID=2325
Uneven mass concentrations and external gravity loads can send satellites wandering. Apollo 16's PSF-2 lunar satellite was only in orbit 35 days. Geosynchronous satellites around Earth have to constantly use stationkeeping fuel to avoid drifting, and the non-spherical shape of Earth is a common consideration for orbital stability in other satellites. MESSENGER was becoming unpredictable; NASA wanted to crash it in a controlled fashion where they had the final say on its landing site for future science.