Again in the same year, on the 28th of May, the house of Mr. Tomes, of Caterham, was struck by lightning, and some slight damage was done. After a careful examination it was found that the greater part of the discharge left the lightning conductor with which the house was provided, and passed over the slope of the roof to an attic room, into which it forced its way through a brick wall, and reached a small iron cistern. This cistern was connected by an iron pipe of considerable dimensions with two pumps in the basement story; and through them the lightning found an easy passage to the earth, and did but little harm on its way. When the earth contact of the lightning conductor was examined, it was discovered that the end of the rod was simply stuck into a dry chalky soil to a depth of about twelve inches. Thus in this case, as in the two former, it was made quite clear that the lightning conductor failed to fulfill its functions because the earth contact was bad.[33] Cases are not uncommon in which builders provide underground a carefully constructed reservoir of water, into which the lower end of the lightning rod is introduced. The idea seems to prevail that a reservoir of water constitutes a good earth contact; and this is quite true of a natural reservoir, such as a lake, where the water is in contact with moist earth over a considerable area. But an artificial reservoir may have quite an opposite character, and practically insulate the lightning conductor from the earth. One which came under my notice lately, in the neighborhood of this city, consists of a large earthenware pipe set on end in a bed of cement, and kept half full of water. Now, the earthenware pipe is a good insulator, and so is the bed of cement in which it rests; and the whole arrangement is identical, in all essential features, with the apparatus of Professor Richman, in which he introduced his lightning rod into a glass bottle, and by which he lost his life a hundred and thirty years ago.