The Mpemba effect is named after Erasto Mpemba, a student in Tanzania who, in 1963, observed that sometimes hot liquid would freeze more quickly than cold liquid. This effect was noted by many scientists throughout history, at least as far back as Aristotle.
The idea that something hot can freeze more quickly than something that is cold sounds counterintuitive. How could it happen?
The problem is that the conditions where this phenomenon might be observed are not well known. How do you compare the two liquids in the first place? Do you measure out equal amounts before heating? Or do you measure the heated liquid to have the same volume as the cold? For that matter, how do you even know when one of the liquids has frozen?
There could be all sorts of reasons why this might happen. If you measured out equal amounts of liquid, and then heated one of the containers, that liquid would expand. That makes its surface area – which determines how quickly heat is lost – greater. So possibly that extra surface area lets the liquid cool more quickly. The objection to this could be that at some point the surface areas of the water in both containers will be equal; so why would one container continue to cool more quickly?
Another possibility is similar, but first you heat the liquid, then measure out the same volume as in the other container. Here, the mass of the hod liquid would be less, because the liquid had expanded. Lower mass might mean the liquid would cool more quickly.
Possibly this effect happens when the containers are both left open to the air. In this case, the heated liquid might lose more heat through evaporation, causing it to cool more quickly.
Ultimately, even though this phenomenon was described as early as Aristotle, we still don’t really know what’s happening.