22 THE TESTING OF KOAD MATERIALS. 



in this test, with a height of fall of 200 cm. The number of blows of 

 the hammer which cause an appreciable breakage is noted. Failure in 

 the test piece is indicated by a smaller rebound of the hammer and a 

 diminution in the height of the specimen. Three dried and three wet 

 specimens are subjected to this test. 



For determining the thrust caused by absorption when expansion is 

 partially prevented, whole paving blocks are used. The block, after 

 thorough desiccation, is placed in a water-tight receptacle and held 

 between two plates at top and bottom, so that the top face of the test 

 piece rests against a block of cast iron, which is stationar}^ The 

 lower plate is supported on the small lever arm of a cement-testing 

 machine. Water at a temperature of 30 C. is poured around the 

 paving block, and as expansion takes place the lever arm of the 

 machine tends to rise. When it rises a valve is opened through which 

 mercury pours into a vessel supported at the end of the arm till the 

 arm conies back to mid-position, when the valve is closed. As the 

 lower plate is fixed when the long lever is in its mid-position, and as 

 this arm is always kept automatically in that place the specimen is 

 prevented from extending longitudinally. The weight of the mercury 

 in the vessel at the end of the short arm is at any moment a measure 

 of the force exerted by the test piece, and as the vessel is supported 

 by a spring balance, this can be read at any, time. Observations are 

 made at first at intervals of three hours then at intervals of twelve 

 hours, and finally at intervals of twenty-four hours till all increase 

 ceases. It will be seen that while this test does not reproduce closely 

 the conditions that exist in practice, the results are of value in giving 

 some indication of the amount of thrust to be expected. After the 

 test the specimens are weighed to ascertain how much water has been 

 absorbed. 



No mechanical tests have yet been devised, or considered necearv. 

 for asphalts. Chemical anah T ses are made, however, to determine the 

 proportion of bitumen, sand, calcium carbonate, clay, pyrites, etc. 



TESTS OF ROAD MATERIALS IN PORTUGAL. " 



The laboratory testing of road materials was tirst taken up in Por- 

 tugal in Ix.M.in the district of Santarem. The IVval test of the 

 French School of Roads and Bridges was adopted, and over >i\ty 

 varieties of rock were toted. The basalts, diorites, and certain of the 

 granites from the valley of the Tagus gave coefficients of wear a> 

 high a> -J-l. This is a higher coeflicient than that of flint, which pre- 

 vious to this test wa> thought to be the best macadam rock, but subse- 

 quent trials on the road have proved this to be a mistake. The 

 coefficients of the limestones tested ranged from 2 to 17. Those from 





aJornal <!"< Kn^cnlirin.s Civilrs IWtuirm'/rs. lssi>, 13: 150. Revista dasObras 

 Publicas e Minas, vols. 13 and L'4; also special ministerial report. 



