268 



Journal of Agricultural Research 



Vol. X. No. s 



in figure i , in which the loss in weight of the various specimens is plotted 

 on a Feret triangle at the points indicating the proportions of coarse, 

 medium, and fine sand in the mix. Several interesting facts are shown 

 in this figure. First, it is seen that the loss in weight of all specimens 

 is practically constant for all mixtures containing more than lo per 

 cent of the coarse sand. This loss, averaging about 5 gm., is very 

 slightly more than would be obtained on a specimen of solid quartz 

 tested in the same way. In other words, the hardness of all specimens 

 containing an appreciable amount of sand over 20 mesh in size appears 

 to be a function of the hardness of the aggregate itself independent of 

 the grading. On referring to the specimens containing various per- 

 centages of medium and fine sand, but no coarse, it will be noted 

 that the loss in weight is very much greater, amounting in two cases 

 ^ to more than 250 



gm. By assuming 

 that the actual hard- 

 ness of the coarse 

 and medium sand 

 grains is the same, 

 in view of the fact 

 that the sand was 

 composed of practi- 

 cally pure quartz, 

 these results show 

 that the use of the 

 fine sand has so 

 weakened the adhe- 

 sion between the 

 cement and the in- 

 dividual particles of 

 aggregate by enor- 

 mously increasing its 

 superficial area that the specimens disintegrated readily under the action 

 of the abrasive. 



In order to determine the hardness of mortar containing very fine 

 sand in a rich mix, as compared to the i-to-2^ mixes, a few tests were 

 made using sands artificially graded as above in the proportion of i 

 part of cement to i X parts of sand. In the case of mixtures containing 

 coarse sand, losses practically identical with those shown on figure i 

 were found. In the case of the very fine sands, however, an average 

 loss of about 40 instead of 200 was noted. This would indicate that the 

 hardness of a mortar containing a large amount of extremely fine sand 

 may be increased considerably by increasing the quantity of cement. 

 Even in this case, however, the average loss is greatly in excess of that 

 obtained by the use of a coarse sand. 



PER CtNT MATe.«lAt_ 



Fig. I. — Graph of hardness tests of artificially graded sand mortar. 

 Proportions i to 2% by weight. 



