306 EEPORT — 1889. 



masonry or cement slaould be formed, which for a one-story building 

 must be -70 m. thick, and for a tvpo-story building 1-20 m. thick. This 

 platform must extend from 1 to 1-50 m. beyond the base of the building. 

 In Manila it is stipulated that the foundations must be able to bear at least 

 twice the weight that is to be placed upon them. When the soil is bad it 

 must be piled or consohdated by a bed of hydraulic concrete, and the 

 foundation of a building must as far as possible be made continuous. 



Another method of minimising the quantity of motion received by a 

 building is to give it free foundations. As an example of this I may 

 mention a room attached to my house which rests at each of its pillar-like 

 foundations upon a layer of ^-inch cast-iron shot placed between two iron 

 plates. Short rollers placed at right angles might be equally effective. 

 This building has stood for many years. It has not been disturbed by 

 typhoons, and at the time of an earthquake a seismograph inside the 

 building shows, relatively to outside, but little motion. Cast-iron balls or 

 shot, even if they are only 1 inch in diameter, cannot be used. They 

 are wanting in frictional resistance, and the building is therefore subject 

 to movements produced by winds and other causes. I do not bring 

 forward this building as an example to be followed in practice, but only 

 as an illustration of a principle which may have practical applications. 

 The first to propose the use of the ball-joint was Mr. David Stephenson, 

 who employed it in connection with lighthouses in Japan. The form he 

 designed was found not to be practical, and is therefore no longer used. 



The ordinary Japanese dwelling-house rests loosely on the upper 

 surface of boulders or stones planted in the soil, and therefore it is difficult 

 to conceive how it can receive the whole of the motion imparted by the 

 shaking ground to its stone foundations. In temples and other large 

 buildings with heavy roofs, which are so common in the country, beneath 

 the supporting timbers and the superstructures there is usually a multi- 

 plicity of timber joints, which at the time of an earthquake yield, and 

 therefore do not communicate the whole of the motion from below to the 

 parts above. In the great earthquake of Ansei, 1855, so far as I am aware, 

 the whole of these buildings remained intact. Certain roofs which are 

 of considerable span in the Engineering College at Toranomon in Tokio 

 were built so that they rested freely on the supporting walls, the object 

 being that they might remain as far as possible resting freely on the 

 walls which moved beneath them. Although they have experienced 

 many tolerably severe shakings, hitherto they have remained uninjured. 

 These examples show, especially for horizontal components of motion, 

 that if a small building is not firmly attached to its foundations, or that 

 parts of a building have connections between them that readily yield, it is 

 difficult to cause such a building to move or swing, and that by a proper 

 application of this principle destruction may be and has been avoided. 



Loose foundations might possibly be employed for small light buildings 

 erected on soft ground. 



For ordinary dwelling-houses, and especially for heavy structures 

 covering a considerable area, I am inclined to the opinion that the solid 

 continuous foundation on the hardest ground, and if possible surrounded 

 by a free area, is the best. The objection to the loose foundations for a 

 large building is, that different parts of the same building do not 

 simultaneously receive momentum in the same direction, and also in large 

 earthquakes there is an actual wave-like motion of the ground. 



