One of the safest houses for an earthquake country would probably be a one-storied strongly framed timber house, with a light flattish roof made of shingles or sheet-iron, the whole resting on a quantity of small cast-iron balls carried on flat plates bedded in the foundations. The chimneys might be made of sheet-iron carried through holes free of the roof. The ornamentation ought to be of light materials. At the time of severe earthquakes many persons seek refuge from their houses by leaving them. In this case accidents frequently happen from the falling of bricks and tiles. Others rush to the doorways and stand beneath the lintels. Persons with whom the author has conversed have suggested that strongly constructed tables and bedsteads in their rooms would give protection. To see persons darting beneath tables and bedsteads would undoubtedly give rise to humiliating and ludicrous exhibitions. This latter idea is not without a value, and most certainly, if applied in houses of the type described, would be valuable. The great danger of fire may partially be obviated by: the use of ‘earthquake lamps,’ which are so constructed that before they overturn they are extinguished. It is said that in South America some of the inhabitants are ready at any moment to seek refuge in the streets, and they have coats prepared, stocked with provisions and; other necessaries, which, if occasion demands, will enable them to spend the night in the open air. These coats, called ‘earthquake coats,’ might also, with properly constructed houses, be rendered unnecessary. _Destruction due to the nature of the underlying rocks._—That the nature of the ground on which a building stands is intimately related with the severity of the blow it receives is a fact which has often been demonstrated. One cause of destruction is due to placing a building on foundations which are capable of receiving the full effects of a shock, and transmitting it to the buildings standing on them. For instance, the reason why a soft bed might possibly make a good foundation, is, as has been pointed out by Messrs. Perry and Ayrton, because the time of transmission of momentum is increased; in fact, the soft bed is very like a piece of wood interposed between a nail and the blows of a hammer—it lengthens the duration of impact. For this reason we are told that a quaking bog will make a good foundation. When a shock enters loose materials its waves will be more crowded, and it is possible that a line of buildings may rest on more than one wave during a shock. There are many examples on record of the stability of buildings which rested on beds of particular material at the time of destructive earthquakes. As the observations which have been made by various writers on this subject appear to point in a contrary direction, I give the following examples:—