50 BULLETIN 124, U. S. DEPARTMENT OF AGKICULTUEE. 



back fill in lci5'ers not over 8 inches thick and thoroughly tamping 

 each layer with a hand tamper. 



The thickness of the successive layers and the weight and face of 

 the tamper should be adjusted to produce the most uniform and solid 

 compression reasonably possible. Conunon practice often is inade- 

 quate, sometimes the layers being made a full 8 inches thick and a 

 light one-man tamper being used. On loose and friable soils a broad 

 face is to be preferred, say 8 by 8 inches, and the weight may be raised 

 to 50 pounds if double handles are provided and the work is done by 

 two men. 



Where the soil is essentially clay and the trench is deep, the back 

 fill should be deposited preferably in courses and a tamper having 

 a smaller face and weighing 10 to 15 pounds used. 



If a clay or silt loam that readily breaks down in water is used, the 

 trench should be filled at first to a depth of not more than 2 or 3 feet, 

 the material being tamped thoroughly as above described, and the 

 partial back fill then may be kept covered with water for several 

 days until the material has become thoroughly saturated and has 

 finished settling. Subsequent layers should be added and " puddled " 

 until the trench is filled completely. Then it should be permitted to 

 stand for several weeks, until the back fill has dried out, before the 

 subgrade is prepared and the pavement constructed. Not infre- 

 quently it is cheaper to haul in gravel, or other material which may 

 be compacted readil}^ for use in back fills, rather than to replace clay 

 according to the method above described. In some cases, where con- 

 ditions were unusually bad, concrete has been used as a back-filling 

 material. Plate VI shows a characteristic failure of a pavement 

 over a culvert where the back fill had been poorly made.' 



The exclusion of vegetable matter from embankments is nearly 

 always provided for in specifications covering road-improvement 

 work, but in a great many cases the importance of this detail does 

 not seem to be realized. As a rule, no especial harm results to a road 

 from vegetable matter being placed in the slopes of embankments, 

 because unequal settlement of the slopes would not necessarily injure 

 the road surface. When sod and roots are placed in the subgrade, 

 however, even in very small quantities, serious damage to the sur- 

 face may result from the porosity caused by the eventual decay and 

 shrinkage of the vegetable matter. Further, roots near the edges of 

 the surface are very likely to live and grow, and the result is a weak- 

 ening of subgrade on a line where the pavement is especially ex- 

 posed to damage. 



Wet-weather springs are not always in evidence at the time a road 

 is being graded, and it is therefore important that the road be exam- 

 ined carefully during wet seasons in order to locate troubles of this 

 kind. Such springs usually may be drained successfully by means of 



