322 



Journal of Agricultural Research 



Vol. X. No. 7 



A rational volume proportion of 88 per cent of aggregate to 12 per 

 cent of bitumen was adopted, and the weight proportions for each com- 

 bination of materials was figured on the basis of this volume ratio. 



All the cylinders in the first series of tests were broken at the end of 

 24 hours, and the results are given in Table IV. Where zero toughness 

 is indicated, the cylinder was deformed by the mere weight of the plung- 

 er's resting upon it. The results in Table IV are tabulated in the order 

 of the increasing strength of the cylinders containing the lightest water- 

 gas tar, and, for convenience, the physical tests of the rocks in Table I 

 are given in the same order. The difference in the relative binding value 

 of the different mixtures is clearly apparent, and it also is definitely 

 shown that certain rocks yield a relatively tough mixture in combina- 

 tion with what generally would be considered a rather fluid tar for bitu- 

 minous construction. However, it will be noted that there is no single 

 physical property which appears to be responsible for this difference in 

 behavior when the rocks are combined with a bituminous material. 



Table IV. — Toughness tests on bituminous-aggregate cylinders 



No. 



6112 



5589 

 1817 

 7316 

 7682 

 7445 

 4813 

 7332 

 6710 

 4444 

 1820 

 871 



Rock. 



Biotite gneiss 



Quartzite 



Metamorphic sandstone 



Diabase 



Biotite gneiss 



Sandstone 



Open-hearth slag 



Limestone 



Blast-furnace slag 



Basalt 



Chlorite gneiss 



do 



Toughness. 



Refined 



water-gas 

 tar. 



Refined 



water-gas 



tar. 



Residual 

 petroleum. 



7 

 10 



13 



14 



9 



12 



14 

 17 

 II 

 12 

 15 

 17 



Oil asphalt, 



Fluxed. 



native 



asphalt 



to 



12 

 12 

 13 

 9 

 13 

 10 



14 

 10 

 II 



14 

 15 



Upon the disclosure of these interesting differences, for which no expla- 

 nation was apparent, a second series of tests was decided upon in which 

 several possible sources of error could be avoided. The crushed material, 

 for instance, was obtained directly from the block out of which had been 

 cut the sections for specific gravity and absorption and the core pieces 

 for hardness and toughness, whereas in the first series the crushed rock 

 was taken from that remaining in a sack after the regular laboratory 

 tests had been made. This change in method tied up the physical tests 

 directly with the material used in preparing the bituminous aggregates. 

 In order to reduce to some extent the possibility of errors due to varia- 



