The question regarding the _interior_ accommodation of the tower must, in like manner, be answered by the nature of the locality where it is erected. Where it is easily accessible, and its stores can be replenished with ease at very short intervals, obviously the interior accommodation may be reduced within very narrow limits. But in exposed situations, as, for instance, on an isolated rock, whose communication with the mainland may be cut off for weeks at a time, room must be provided for ample supplies, and conveniences for the keepers must be arranged on a liberal scale. In the long and dreary nights of winter, where, in the northern parts of Great Britain, it is necessary to keep the light burning for about seventeen hours, not even for a moment is it left without the watchful care of at least one keeper; and thus, as he will require an interval of repose, its superintendence will occupy _two_ persons; but in open, exposed places like the Eddystone, the Bell Rock, the Wolf, and the Skerryvore, where it is frequently impossible to communicate with the mainland for three, four, and even six weeks, circumstances have rendered it desirable that there should not be fewer than _three_ men on duty. Hence, sleeping apartments have to be provided, as well as receptacles for sufficient supplies of water, food, fuel, and other matters. * * * * * The second condition to which we have adverted is, that the building shall be capable of resisting the force of the wind and waves. The wind is baffled with comparative ease, but the sea is a far more formidable, and, moreover, is an ever-present foe. Even in the summer months the pressure of the waves is very considerable, averaging, perhaps, about 611 lbs. per square foot of surface exposed to it. In the winter, however, the average rises to 2086 lbs. per square foot; while, during stormy weather, the force has amounted to no less than 4335 lbs.[23] To oppose this immense pressure, not only must the masonry be of the solidest description, but such a form must be given to the building as will expose that masonry to the least possible stress. From various experiments it has been found that the most effective form is that of the cylinder; and with certain modifications, the cylindrical is now almost universally adopted in the erection of lighthouses. Scientifically speaking, however, it is not so much a cylinder as the union of frustra of different cones, with a curve osculating the outline of the successive frustra. To the youthful reader this may not be very intelligible, and we will, therefore, refer him to the form of the Eddystone as an illustration of what we mean. Smeaton himself relates that it was suggested to him by the trunk of an oak, but there seems reason to believe that this was an after-thought, intended for the benefit of the large class of minds which cannot appreciate scientific reasonings.