Lightning discharges do not take place between cloud and earth only, but also, and perhaps more frequently, between too oppositely charged clouds. We then get atmospheric lightning, the flash often extending for miles. This form of lightning is harmless, and in all probability what we see is only a reflection of the discharge. The oft-told tale of the lightning flying in at the window, across the room, and out of the door, or up the chimney, is all moonshine, and before dealing with lightning protectors I intend to expose some of the fallacies concerning lightning. Were the discharge to pass through a house, it would infallibly leave more decided traces and do more damage than simply scaring a superstitious old lady now and again. Many people are often and unnecessarily frightened during a thunderstorm, but it may be safely predicted that a person under a roof is infinitely safer than one who is standing alone on level ground, and making himself a prominence inviting a discharge. Rain almost invariably accompanies the discharge, and the roof and sides of the house being wet, they form a more or less perfect channel of escape should a flash strike the building.--_Knowledge._ * * * * * RESEARCHES ON MAGNETISM. By M. DUTER. If we place a thin plate of steel in a uniform magnetic field, so that the lines of force of the field may be normal to the surface of the plate, we have a very flat magnet, the two faces of which are the two polar surfaces. The magnetic distribution thus obtained seems to disappear when the plate is no longer in the field. The following experiments show that this disappearance is not complete. I made use of plates of tempered steel of 1 millimeter in thickness, and varying in diameter from 0.040 to 0.005 meter. With these plates I formed cylindrical batteries. In some of these batteries the plates are directly in contact, and in others they were separated by leaves of pasteboard, the thickness of which varied from that of the thinnest paper to 0.001 meter. The batteries were placed in the central portion of a very powerful magnetic field, and after they have been taken out they formed perfectly regular permanent magnets. The supporting power of these magnets was the greater the nearer its constituent plates were to each other. In a battery of 100 plates, touching each other directly, and strongly pressed into a brass cylinder, the portative force at each extremity rose to 30 grammes. This first result having been obtained, I dismounted the batteries, plate by plate, taking care to mark the upper and under side of each. I found then that each plate retained only an excessively slight magnetism. Yet each of them still constituted a flat magnet, of which the two faces are the polar surfaces; for on rebuilding the battery it gave again a perfectly regular magnet, though weaker than it was at first. The separation of the magnet