456 ADAM.S. 



and comparisons made with the results obtained in the AYashington 

 Laboratory, wliic-h has been in operation for a number of years. 



The Devall abrasion machine consists essentially of two iron cj'linilers fastened 

 to a sliaft so that each cylinder is at an angle of 30° with an axis of rotation. 

 Eacli cylinder is 20 centimeters in diameter and 34 centimeters in depth, closed 

 at one end and having a tightly fitting iron cover for the other. In testing stone 

 with this machine it is broken into pieces of as nearly uniform size as possible, 

 each of which will pass in all positions through a centimeter ring. Five 

 kilograms approximating as closely as possible to 50 pieces of rock previously 

 cleaned by washing and subsequently dried in an oven and then cooled in a 

 desiccator, is used as a charge for each cylinder. The machine is rotated at the 

 rate of 2,000 revolutions per hour for five hours, the inclination of the cylinders 

 causing it to be thrown from one end to the other twice at each revolution. When 

 the abrasion of the material is completed, the contents of each cylinder is placed 

 on a sieve having openings of 0.16 centimeter, and the material which passes 

 through this is caught on a sieve having openings of 0.025 centimeter. Both 

 sieves with the rock held on them are placed under running water until all the 

 dust is washed ofl'. These fragments are then dried in a hot air bath and cooled 

 in a desiccator and weighed. 



Percentage of wear. — The percentage of weaj- indicates the amount of 

 the original sample "which was worn to fines less than 0.16 centimeter in 

 size during the test. Theoretically, this percentage of wear may vary 

 from to 100, but materials which show 20 per cent or more of wear 

 are considered unfit for use. Some materials will show less than 3 per 

 cent of wear. 



French, coefficient. — Tlie French coefficient, wliich is an arbitrary one, is 

 determined by dividing 2,000 by the weight of the fragments under 0.16 

 centimeter in size, which are worn off during the test. According to this 

 standard, the scale of excellence would vary from O^to 30, with some 

 material passing 30. 



United States coefficient. — In reports of the road material laboratory 

 at Washington, the United States coefficient is also given, which is found 

 by dividing by 10 the weight of the material over 3 centimeters in size 

 which remains after the test. 



Report of a series of tests. — Inasmuch as the French coefficient is an 

 arbitrary standard and the United States coefficient does not express so 

 directl}' the wear in actual use on the roads, it is not considered neces- 

 sary to record them, although they have been computed in all the tests 

 made in this Bureau. The following table is a summary of the abrasion 

 tests thus far made on road materials used in the Philippines. It is 

 presented for the information of those who are in charge of the con- 

 struction of roads with tlie hope that it may indicate the wide range 

 in the character of the stone used and show that in many cases a con- 

 siderable economy may be effected by carefully selecting the quarries. 



