To: Cc: Subject: ice spikes The explanation from Henry Bloomfield is basically correct but it's missing a few details. Although the rate of freezing has something to do with the spikes, it does not have to happen very slowly, just not very fast. The type of spikes you have are specific to the fact that you are using a plastic ice cube tray and it's the pressure from the plastic tray that forces the water to rise up out of the a small hole as the cube freezes. Metal trays will not produce the spikes. Different spikes can also form under different conditions as this phenomenon has been observed in frozen puddles and plastic flower pots with frozen water outdoors. There have been some that are a half a foot high reported on the mountain tops in hawaii. The spikes you see in your freezer should typically have a triangular cross-section, with an occasional rectangular cross-section, unless they have become rounded after melting. The angle and cross-section of the spike is determined by frozen sheets of ice forming within the ice cube. These ice sheets actually begin to crystallize within the ice cube before the top surface or side walls of the cube begin to freeze. If you time it right, about an hour and a half to two hours after putting in a fresh ice tray, you can check the cubes to see that ice sheets within the cube begin to form before anything else begins to freeze. These ice sheets form random planes that are suspended within the cube and must not be disturbed for spikes to form. That is why there is better chance of spikes forming when the freezer is not disturbed. If three of these sheets of ice should form a triangle, they have the ability to form a funnel. Once this funnel is formed and is near the surface, the surface can lock the funnel in place by freezing. If the conditions are not right, then the funnel will close and form a triangular divot on the surface of the ice cube. If the conditions are right, this funnel will maintain a liquid center and provide the most efficient path for the latent heat of the cube to escape during the freezing process. The shape of this funnel is usually triangular, as that is the most likely configuration. The shape of the funnel beneath the frozen surface of the cube is also what determines the angle of the spike. If the funnel is vertical, then the spike will be vertical, but if the funnel is slanted, then the spike will form with an angle as it is pushed out of the hole. Once the surface of the cube is completely frozen leaving a triangular hole, the pressure from the plastic tray as the ice expands will force the water within the cube that is still liquid to come out of the hole. If the rate of the water coming out of the hole is close to the rate of ice freezing, then the edges of the triangular funnel will freeze leaving the center liquid to bulge upwards. The edges of the bulging water then freezes at a rate that conicides with the rising water. If the freezing is too fast, then the bulges freezes over and forms a little pyramid shape. When the freezing rate matches that of the rising water, a tube is formed to allow the water to flow upwards until it runs out of liquid water or eventually tapers off and the tube closes due to freezing. If you can catch a spike during formation before it closes off, you'll be able to see that the spike is formed out of a hollow tube and as soon as you open the freezer and remove the tray, the water level inside to tube will begin to drop due to a decrease in the rate of freezing. A fresh ice cube spike will have a completely triangular cross-section and will only be rounded due to melting of the top portion. Hope you're still getting these ice cube spikes. Btw, nice observatory. I teach intro astro/physics at uc berkeley. miles chen