...how...
The electromagnet is easy to explain, faulty wiring can easily break one. Heck, even non-faulty wiring that is arranged wrong can (simply have the current run an equal number of times in the opposite direction, and you have an equal magnetic field with an opposite direction, effectively canceling the field, though it would be a dipole-ish-thing).
The....light....how? Only thing I can think of is that it was mislabeled...or miscalculated, or the error was very large (well, relatively). What experiment was that one?