16 CEMENTING POWER OF ROAD MATERIALS. 



which experiments are now being made. Clay itself, as every one 

 knows, is generally too soft and plastic to bear traffic in wet weathev, 

 although its binding power is high. By burning and clinkering a 

 portion of the clay to be used on the road its hardness is increased and 

 its binding power destroyed. By proper constructive methods and 

 mixing we approach the conditions obtained with rock dust and with 

 sand-clay mixtures. 



It may be objected that at least one exception to the general rule is 

 furnished by certain furnace slags which certainty have not undergone 

 secondary changes, and yet have been successfully used on roads in 

 some localities. This subject has been especially investigated, and it 

 has been found that siliceous slags are in no case suitable for a binder 

 surface. Certain basic slags, which have an excess of lime in their 

 composition, and which have been quickly cooled at the furnace, make 

 an excellent material for light-traffic roads, as the slag dust takes on 

 an actual mild set similar to that of Portland cement. Doughs made 

 of such slag dust set hard before they can be aged twenty-four hours 

 preparatory to molding the test pieces. It is apparent that this special 

 case of slags has no general bearing upon the conclusions reached. 



It happens that certain ferruginous limestones and gravels bind 

 exceedingty well, and this has led to an erroneous impression, which 

 prevails even among engineers of long experience, that "iron' 1 is the 

 cause of the bond of cementation. Broadly speaking this is not true, 

 for many of the highly ferruginous red sandstones do not bind. A 

 series of experiments was made in the laboratoiy to determine 

 whether iron ores, such as hematite and the hydrated oxid, limonite, 

 possessed cementing power, and it was found that exceedingly pure 

 limonite had a low value and hematite none at all. If, however, there 

 were siliceous or clayey impurities present, the values would be much 

 higher. It has been frequently observed that when metallic iron 

 tilings as, for instance, in the neighborhood of a machine shop are 

 thrown upon a footpath, the path consolidates into a hard impervious 

 shell under the light impact of human feet. This bond may be due to 

 the mere rusting together of the iron particles, or it is possible that 

 hydroxid of iron, which is formed under the influence of water and 

 oxygen, is colloid in its nature, and thus acts as an ordinary binder. 

 If hydroxid of iron is precipitated by the wet "method in the labora- 

 tory and the resulting colloid precipitate after partial dn r ing out is 

 mixed with inert material like tine sand or ground glass, the hydroxid 

 on drying will bind the particles together. In the same way it is pos- 

 sible to prepare in the laboratory colloid preparations of alumina, 

 silica, and various silicates, such as those of aluminum, calcium, iron, 

 and magnesium. All these substances appear in rock structures, 

 gravels, clays, and soils, and all of them were found to act to a greater 

 or less degree as binders when tested in the laboratory. 



