788 EXPERIMENT STATION RECORD. [Vol.41 



point in the deposit can be regarded as representing witliin the i^robable error 

 due to laboratory manipulation the percentage of wear of the stone over quite 

 a considerable area, at least a quarter of a square mile. It is possible to 

 assign average values and fairly narrow limits of variation of this value for 

 the percentage of wear and the toughness of the material occurring in a lime- 

 stone formation covering areas up to 50 and 60 square miles and with thick- 

 nesses up to 500 ft., even though stone varying in character is included in the 

 formation. These values and limits will include the majority of results of 

 tests made on samples collected throughout the formation and its various 

 horizons. . . . The results of abrasion and toughness tests over a series of 

 diabase dikes of the same age and structure reveal very uniform results in 

 deposits of the same grain. A change in the average size of gi-ain affects the 

 toughness value slightly." 



In the investigation on bowlder aggregates it was found that bowlder de- 

 posits in any one area consist of many combinations of three or four rock 

 types of fairly uniform durability, and that the percentage of wear of any 

 combination of the rock types found in deposits of bowlder aggregates could 

 be calculated by the following formula if the percentages of wear of each of 

 the rock types were known : 



w ^^^ 

 '"— 100 ' 



in which TFi, TT, TFot^^ percentages of wear of the various rock types, and Cj, 



C2 Cn=percentage proportions in which the rock types are present in the com- 

 bination. 



In the sand and gravel investigation a large variation was found between 

 the results of duplicate granulometric laboratory analyses on the same sack of 

 gravel. The variation in texture over one deposit of gravel of 800 acres was 

 found to be large. 



An abrasion test for stone, gravel, and similar aggregates, H. H. Scofield 

 (Amer. Soc. Testing Materials Proc, 18 (1918), pp. 416-^28, figs. 7).— It has 

 been found that the standard Deval test for road materials is somewhat mis- 

 leading in its results due to the retention within the abrasion chamber of the 

 dust worn from the charge. An apparatus is described which was devised 

 at Purdue University, to fill the need of a rapid and practical abrasion test 

 for road materials which is simple in construction and allows the dust of 

 abrasion to escape. The machine is a small type rattler in which dust and 

 chips from abrasion escape between the staves as fast as formed. The open- 

 ing between staves is ^s in. The abrasion chamber is octagonal in shape with 

 a volume equivalent to that of the Deval cylinder. Recent tests with this appa- 

 ratus upon 12 Indiana limestones show in general that the new test gives a 

 greater range of results and a consequent better differentiation of quality. 



Effect of controllable variables on the toughness test for rock, F. H. Jack- 

 son, JR. (Amer. Soc. Testing Materials Proc., 17 (1917), pp. 571-588). — Studies 

 on the effect of certain controllable variables on the accuracy of the tough- 

 ness test for rock for road building are reported. 



It was found that " the quality of the product of a rock quarry may vary 

 from time to time to such an extent that it is advisable to test the material, 

 as nearly as possible, at the time it is to be used. Great care should be exer- 

 cised in selecting samples for the toughness test in order to insure obtaining an 

 average value for the toughness of the entire product. All materials exhibiting 

 any indication of foliation or bedding should be tested in two directions; one 

 set perpendicular and the other set parallel to the foliation.s. A greater total 

 variation than 1 mm. in either the diameter or height of the test specimen 

 should not be allowed. Variations in the moisture content of test specimens 



