FORMS FOE SPECIFICATIONS, ETC., FOR ROAD MATERIALS. 35 



uniformity of mixing of the three usual ingredients and upon the adequate 

 consolidation of the mass. From the practical standpoint the selection of the 

 raw material should not be based on a too narrow analytic figure. The natural 

 deposits are of variable composition. A number of samples from a selected 

 area close to the work may not conform to a specific percentage, but it is 

 usually possible to supplement deficiencies by adding relatively small quan- 

 tities of sand or clay, as the case may be, and secure in the mixed roadbed a 

 fairly uniform material whose analysis is satisfactory. 



The final acceptance of a material of this class should preferably be based 

 on samples taken at specified intervals on the freshly laid and intermixed 

 roadbed and at such other points as the inspector may choose where the indica- 

 tions are that the mixture is not uniform and satisfactory. 



In the light of the above brief statement the specifications, methods of test, 

 methods of sampling, and form of reporting test results on these materials 

 will be more readily understood. 



(12) Examination Semi-Gravel, Top-Soil, and Sand-Clay: 



Dry 500 grams of the material at a temperature below 350° F. (176.6° O.) 

 to constant weight. Gently pulverize to break down soft clods or masses, 

 but not to grind or break hard material. Pass through a 10-mesh sieve, weigh 

 the coarse residue and record as " coarse material." Use the material passing 

 the 10-mesh sieve as the starting point of a percentage analysis -as follows : 



Weigh out two samples of 50 grams of this material for duplicate analysis. 

 Place each in a wide mouth bottle (5 to 6 cm. diameter and about 12 to 15 cm. 

 high). Add about 5 c. c. of dilute ammonia water and about 200 c. c. of water. 

 Close with a cork or glass stopper and shake thoroughly for 20 minutes. Allow 

 the sample to settle eight minutes and decant carefully or siphon off the super- 

 natant liquid to a depth of 8 cm. below the surface of the liquid. (The depth 

 of the liquid in the bottle should be sufficient to leave about 4 cm. below the 

 point of siphoning.) Fill the bottle again with water, shake for three minutes, 

 allow settlement, and siphon off as before. Repeat the process until the super- 

 natant liquid is clear. Be careful to wash the stopper and neck of the bottle 

 free from coarse material before decanting. The washings drawn off may be 

 collected and evaporated to dryness for direct recovery of the fine sediment 

 classed as " clay." Ordinarily the sediment loss or clay is computed as dif- 

 ference. 



Wash the contents of the bottle cleanly into a porcelain evaporating dish and 

 carry to dryness on a water bath. The dried residue should be carefully scraped 

 from the dish and passed through a nest of 20, 60, 100, and 200 mesh sieves. 

 The residue retained on each sieve is weighed and recorded as sand of the 

 respective sizes. Their sum constitutes the total " sand." The residue passing 

 the 200-mesh sieve and caught in the pan is weighed and recorded as " silt." 

 Duplicate samples should check within 1 per cent. 



(a) The coarse material should be examined for hardness and with the mag- 

 nifying glass to identify its character as quartz, hard iron compounds, feldspar, 

 schistose material, or indurated clay. Hard quartz or iron gravels are valuable 

 in themselves and as indicating the quality of the finer aggregate. Feldspar, 

 mica, and clay nodules are worthless and indicate that the accompanying soil 

 is poor for road building. 



(6) The sands should be examined with the magnifying glass for identifica- 

 tion as quartz and for the presence of mica scales or feldspar needles. If mica 

 or feldspar is present in appreciable amounts the sample should be rejected. 



(c) When the clay is recovered by evaporation it can be examined for tenacity 

 by cementing together two glass plates, each 1 inch wide, set at right angles, 

 with a layer of clay whose thickness is fixed by a fine bent wire laid between 



