272 



Journal of Agricultural Research 



Vol. X, No. s 



f /\'/\/\J\ A A /V 

 7\ /\ /\ AA A 



A /\ /\^/\ /\ A-'A'A 



180 IJ5 laz 200 t zoa jij 



./\./\./\/\x\./\ 



Fig. 7. — Graph of tension tests of artificially graded sand mortars. 

 Proportions i to 214 by weight. 



and each value in compression the average of three tests. The average 



total variation in individual results of each set was lo per cent for tension 



and 5 per cent for com- 

 pression. The results 

 show graphically the 

 enormous variation 

 which may be obtained 

 in both the tensile and 

 compressive strength of 

 mortar by reason of vari- 

 ations in the grading of 

 the sand. The contour 

 lines are much more 

 regular than those shovv^- 

 ing variations in tough- 

 ness. The following 

 points may be noted : 



Tension. — Sand s 

 showing maximum re- 

 sults in tension are com- 

 posed of from 60 to 80 



per cent coarse, o to 20 per cent medium, and 10 to 30 per cent fine, 



whereas the weakest sands are those composed of from o to 10 per cent 



coarse, o to 20 per cent .^^ 



medium, and 80 to 100 



per cent fine. A sand 



composed of 100 per 



cent coarse is 35 per 



cent stronger than one 



composed of medium 



only and 80 per cent 



stronger than one com- 

 posed of fine only, but 



is nearly 40 per cent 



weaker than the sand 



of maximum strength 



(70-10-20). The total 



variation in strength is 



250 pounds per square 



inch, or 54 per cent 



from the average of 



the 60 determinations. 



Compression. — Sands showing maximum results in compression are 



composed of from 60 to 80 per cent coarse, o to 20 per cent medium and 



10 to 30 per cent fine, the same limits as for tension. A sand composed 



AyyTyTYTT 

 /'•^ 



/ A A 



A A\ A /AAA A A 



\ 



PER Ctt»T MATCRIAL UHOLK NO 50 Sitvc. 



Fig. 8. — Graph of compression tests of artificially graded sand mcrtars. 

 Proportions i to 2/4 by weight. 



