A good lightning conductor should not have a higher electrical resistance than 10 ohms from the point to the ground, including the "earth" contact. Exceptionally good conductors have only about 5 ohms. A high resistance in the rod is due either to a flaw in the conductor or a bad earth connection, and in such a case the rod may be a source of danger instead of security, since the discharge is apt to find its way through some part of the building to the ground, rather than entirely by the rod. It is, therefore, important to test lightning conductors from time to time, and the magneto-electric tester of Siemens, which we illustrate in figures 98 and 99, is very serviceable for the purpose, and requires no battery. The apparatus consists of a magneto-electric machine AT, which generates the testing current by turning a handle, and a Wheatstone bridge. The latter comprises a ring of German silver wire, forming two branches. A contact lever P moves over the ring, and is used as a battery key. A small galvanometer G shows the indications of the testing current. A brass sliding piece S puts the galvanometer needle in and out of action. There are also several connecting terminals, b b', l, &c., and a comparison resistance R (figure 98). A small key K is fixed to the terminal l (figure 99), and used to put the current on the lightning-rod, or take it off at will. A leather bag A at one side of the wooden case (figure 99) holds a double conductor leading wire, which is used for connecting the magneto-electric machine to the bridge. On turning the handle of M the current is generated, and on closing the key K it circulates from the terminals of the machine through the bridge and the lightning-rod joined with the latter. The needle of the galvanometer is deflected by it, until the resistance in the box R is adjusted to balance that in the rod. When this is so, the galvanometer needle remains at rest. In this way the resistance of the rod is told, and any change in it noted. In order to effect the test, it is necessary to have two earth plates, E1 and E2, one (El) that of the rod, and the other (E2) that for connecting to the testing apparatus by the terminal b1 (figure 99). The whole instrument only weighs about 9 lbs. In order to test the "earth" alone, a copper wire should be soldered to the rod at a convenient height above the ground, and terminal screws fitted to it, as shown at T (figure 99), so that instead of joining the whole rod in circuit with the apparatus, only that part from T downwards is connected. The Hon. R. Abercrombie has recently drawn attention to the fact that there are three types of thunderstorm in Great Britain. The first, or squall thunderstorms, are squalls associated with thunder and lightning. They form on the sides of primary cyclones. The second, or commonest thunderstorms, are associated with secondary cyclones, and are rarely accompanied by squalls The third, or line thunderstorms,