CAUSES OF CEMENTING POWEK. 15 



quarried rock. It will be found, however, that the considerations are 

 identical. Clays themselves are the product of rock decay under the 

 action of water and watery solutions. The essential clay base is sup- 

 posed to be a hydrated silicate of aluminum, known as kaolin, which is 

 the result of the action of water on the double silicates of aluminum 

 and the alkali metals. The most typical of the double silicates are the 

 feldspars, a class of minerals very widely distributed in nature and 

 occurring as a crystalline ingredient of a great number of different 

 rock species. The microscopical analysis of rocks shows that a great 

 many of the minerals which make up the aggregate structure have 

 undergone secondaiy changes similar to, if not identical with, kao- 

 linization of feldspar. Now, kaolin, as found in nature, although not 

 lacking in binding power, does not, as a rule, excel in this quality. 

 The ball claj^s are usually added by potters to the purer kaolins and 

 china clays to increase both the plasticity and the binding power. It 

 is very evident that binding power is not due to the presence of a par- 

 ticular mineral such as kaolinite; on the contrary, the higher binding 

 clays show a preponderance of amorphous rather than of crystalline 

 particles. 



The evidence obtained in this laboratory points to the following 

 conclusions: Many minerals under the action of water or of watery 

 solutions are decomposed. The secondary products, which are usually 

 highly hydrated, may or may not lead to binding power, as they are 

 capable of existing in allotrophic modifications which differ in this 

 respect. Alumina and many other substances can be easily prepared 

 by wet reactions in the laboratory, either as gummy colloids or as 

 finely crystalline precipitates, by slightly varying the conditions. In 

 nature the conditions are of every possible variety, and thus we find a 

 physical property like cementing power varying through wide limits 

 in those rock species which exhibit it. In a word, those rock dusts 

 which contain a certain proportion of particles which on soaking with 

 water soften to the extent that they become, to ever so slight a degree, 

 glue-like (colloid), and thus adhere, are those which are useful to the 

 road builder. Many of the traps, limestones and sandstones, fall 

 under this head. Those rocks, on the other hand, which are of an 

 entirely unaltered crystalline structure, or those which, through meta- 

 morphic changes heat and pressure have had the active hydrated 

 particles destroyed such as slates and quartzites should be avoided 

 on the surface portion of the road. 



These conclusions are borne out in service where the problem has 



be solved of building roads of material which, while hard enough 

 to bear traffic, is without binding power. A good example is fur- 

 nished by the excellent sand-clay roads in the Southern States, 

 described in the Yearbook of the Department of Agriculture for 

 L903. A somewhat similar case is that of burnt-clay roads, which 



ive been advocated by this laboratory for several years and upon 



