In the intermolecular forces/states of matter investigation called "Don't Flip Your Lid" students are asked to observe the order in which four different chemical substances melt. The appropriate order in terms of increasing melting point is paraffin, iodine, sugar, salt. Just to note, the salt will not melt at the temperatures we are working with and the sugar will start to burn and smoke after melting. The iodine will sublime and evaporate (typically in that order) and none will be left at the end of the lab.

The procedures call for a bunsen burner, but I have found that the increase in temperature is so sudden and dramatic that students will usually cite iodine as the first to melt. To correct for this, I have my students make little trays out of aluminum foil. By making folds it is easy to create four small sections of the tray that can hold all compounds, keep them separated, and sit flat on a hot plate. I make sure that the students start with the hot plate unplugged. Then they plug it in and crank the heat. Because it is starting from a completely cooled position, the increase in temperature is much steadier and students can see the paraffin pooling as it melts just before the iodine starts to do the same as it simultaneously vaporizes in a beautiful purple haze.

This particular sequence in melting point is what makes the conversation around the data so rich. Despite having less total electrons, and though it is also nonpolar, the iodine actually has a greater melting point than paraffin. This allows you to discuss the idea that polarizability of the electron cloud is affected by both the number of electrons and the "squishiness" of the cloud as it relates to the energy levels those electrons are in. For example, all of the electrons in paraffin are in the second energy level or lower. The polarizability is hence diminished. It would be like trying to shift the stuffing around in a firm futon mattress. On the other hand, the electrons in diatomic iodine are in the fifth energy level and lower. Polarizing the electron cloud is now more analogous to shifting the stuffing around in a fluffy down pillow. More polarizability means greater london dispersion forces and a higher melting point.

So, long story long, switching over to a slowly heating hot plate gives much more accurate data (from my experience at least).