We will be endeavoring to answer these and see if we were really just fooling ourselves.  If we are fooling ourselves, a change in air flow is the most likely culprit.  The emissivity of the wires was most likely unchanged between periods of higher measured temperature, so making opaque wrappings for the glass doesn't seem like a great diagnostic next step.  Instead, we are pursuing stabilizing the thermal environment around the cell, first.  We had previously put the cells in a box made from twin wall polycarbonate and fed the box with constant temperature air.  This time we are using extruded polystyrene foam board, some aluminum, and a couple fans to make a higher performance isothermal box.  Nothing wrong with making an opaque tube wrap at that time, too, I guess.