With the desire to get further light on the subject, I consulted the great experimental physicist, my friend Sir James Dewar, in his laboratory at the Royal Institution. He told me that the late Professor Tyndal used to exhibit the production of flame by the friction of two pieces of quartz in his lectures on heat, but made use of a very large and rough crystal of quartz (rock-crystal) and rubbed its rough surface with another large crystal. Tyndal's note on the subject in his lecture programme was as follows (Juvenile Lectures on Heat, 1877-78): "When very hard substances are rubbed together light is produced as well as heat." Sir James Dewar kindly showed me the crystals used by Tyndal, the larger was 16 inches long and 4 or 5 inches broad. We repeated the experiment in the darkened lecture room, and obtained splendid flashes. The same smell is produced when rock-crystal is used as when flint or quartz pebbles are rubbed together. All three are the same chemical body, namely, silica (oxide of silicon). We also found that when the crystals were bathed with water or (this is a new fact) with absolute alcohol, the same flashing was produced by the friction of one against the other. Later, with the kind assistance of Mr. Herbert Smith, of the mineral department of the Natural History Museum, I examined, with a spectroscope, the flash given by two quartzite pebbles when rubbed together. No distinctive lines or bands were seen; only a "continuous" spectrum, showing that the temperature produced was not high enough to volatilize the silicon. I also examined some pebbles of another very hard substance--nearly as hard as silica (rock-crystal, quartz, and flint). This was what is called "corundum," the massive form of "emery powder" (oxide of aluminium). By grinding two of these corundum pebbles with very great pressure one against the other (using much greater pressure than is needful in the case of quartz), I obtained flashes of light. It was not known previously that any pebbles except those of silica would give flashes of light when rubbed together. A smell resembling that given out by rubbed quartz, but fainter, was observed. Those are the facts--new to me and to many others--about this curious subject. The flashing under water is a very remarkable thing. I cannot say that I am yet satisfied as to the nature of the flash. A simple explanation of the result obtained, when two dry pebbles are rubbed together in the air, is that crushed particles of the quartz or of the corundum are heated by the heavy friction to the glowing point. But this does not accord with the fact that submergence in a liquid does not interfere with the flashing. The rise of temperature would certainly be checked by the liquid. And the curious smell produced is in no way explained.