It is not very easy to give exact prescriptions as to the best manner in which the underground connection should be effected. The means vary entirely with the circumstances, and the matter should in all cases be intrusted to an expert. Simple as is the whole theory of lightning protection, consisting in nothing else but laying a good metallic path from the top of a building down into moist earth, as an unfailing path for the electric force, the practical execution of it is not the less often very complicated. It is especially so as regards the most important of points, that of the underground connection. Of course, wherever there is running water at hand, a river, or even a tiny stream that never dries, the matter is easy enough, but as in the great majority of buildings to be protected such water does not exist, the solution of the question becomes more difficult, and frequently one of the greatest perplexity. It tends even to be more and more so in consequence of the progress of sanitary arrangements under which towns and villages are ‘drained’ until the soil has been made as dry as a rock. Immense as the benefit is to public health, it is, like all benefits, attended by certain drawbacks. One of these certainly is a greater danger from lightning. It is often proposed by builders to use the drain-pipes themselves in making ‘good earth’ for lightning conductors, but the fallacy of this recommendation need scarcely be exposed, seeing that these conduits are generally made of earthenware, as happened when Christchurch, Clevedon, was struck by lightning. While broad rules cannot be laid down, still it may be affirmed that a good earth connection, sufficient to carry off the heaven’s electric discharges, may always be obtained by either of two means. The first, and in all cases most preferable, is to lay the conductor deep enough into the ground to reach permanent moisture. When this exists in a considerable mass, the single conducting rope, touching it, will be quite sufficient; but when the quantity is deficient, or doubtful, it will certainly be advisable to spread out the rope, so as to run in various directions, similar to the root of a tree, likewise in search of moisture. There are various modes of accomplishing this, shown in figs. 46 and 47. [Illustration: Fig. 46.] [Illustration: Fig. 47.] A variety of methods have been proposed for the dispersion of the electric force underground where the soil contains little or no moisture, except at great depths, to be reached only by a vast amount of labour and expenditure. In France, the system most generally adopted in these cases is to place at the bottom of the underground connection an apparatus, made either of iron or copper, shaped somewhat in the form of a harrow, and to embed it thickly in charcoal. Fig. 48 will illustrate this system of earth connection. [Illustration: Fig. 48.]