356 Transactions of the American Institute. 



alluded, more frequently occur from subsequent changes, beneath 

 the superstructure, or from irregularities in the support which the 

 foundations afford at different places. 



A foundation on durable rock may be regarded as perfect, but 

 even in this case, care must be taken to avoid placing a heavy struc- 

 ture upon steep sloping surfaces upon which it may slide, unless 

 it is prevented by benching down the rock in horizontal planes. 



Foundation masonry below ground is usually of a less compact 

 character than that above, and if, as often occurs, the rock is found 

 at a much lower level at one place, than it is at an adjacent one, 

 tlie unequal compacting of such masonry of different thicknesses 

 will injure the works resting upon them. Hard compact earths 

 are ccmsidered good foundations; under heavy pressure they may 

 be somew'hat compressed, but if this is uniform no injury ensues; 

 ■\rhen saturated w^ith water, this compression will be greater, and 

 there is sometimes danger that the saturation will occur, in one 

 place and not adjacent, and thus produce an unequal settlement. 



Gravel and sand also afford good foundations, but are more liable 

 to escape laterally, and allow the structure to settle. 



Loam, silt and other light soils are so compressible that where 

 t'hey occur it is usually necesary to construct an artificial foundation. 



The most difficult case, and one of common occurrence, is to 

 provide a proper support for a heavy structure, when the natural 

 material is not uniform, under all portions of the foundation; and 

 this requires the greatest circumspection on the part of the engineer, 

 and taxes his skill to produce uniformity of support at the least 

 expense.* This is frequently done by expanding the base of the 

 foundation where it is less firm, and that requires a knowledge of 

 the supporting power of each different description of material, 

 which must often be ascertained by direct experiment. 



Rock is incompressible, and some earths, when dry, nearly soj 

 Band and gravel is but slightly compressible when confined laterally, 

 as in deep pits. Good soil will sustain from five to ten tons per 

 square foot, but these weights will compress it a little. Earth, 

 saturated with water, under heavy loads, will not only be com- 

 pressed, but will also escape laterally whenever the frictional 



• In a recent letter from John B. Jervis, to the author, he says: "I have read with 

 satisfaction your riews on 'Foundations,' the most difficult as well as the most important 

 part of tie science of engineering. To provide a support for a structure that shall be 

 adequate, and with no great surplus, is a far more difficult problem than to make it sure at 

 a much greater expense." 



