I've been playing in the lab with distilled water and noticed that the smaller water drops are, the longer I could over cool them. In fact, even with 20 ml of distilled water, I couldn't get it to freeze when it passed by 0.0°C. Instead it would go around -3.5°C, then freeze almost instantly and the temperature would also rise almost instantly up to 0.0°C. With small drops I could reach easily a temperature of -8 to -13 °C before they would freeze. And with very very small drops, they could reach around -15°C or so. I'd have thought that the smaller the sample of water, the faster it would freeze but the opposite occurred. 1)Why is it so? I've read on the Internet that an argument used to explain Mpemba's effect is that boiled water will evaporate faster than "cold water" (this is true), but then they say that this imply that it will freeze faster. According to what I've done in the lab, less water would imply slower freeze due to a longer over cooling. 2) So are their argument flawed? I tried to get rid of over cooling by vibrating the freezer, I could see (thanks to a camera inside the freezer) that all the surface of the drops was affected by my vibrations but it would not get rid of the over cooling... So basically my first question is equivalent to ask why is there a greater over cooling when the water sample is smaller. Thank you.