20 BULLETIN 1216, U. S. DEPARTMENT OF AGRICULTURE. 
Example : 
Assume — Base mix, 1 : 2 : 34 i».\ volume. 
Unit weights — 
Fine aggregate, 95 pounds per cubic* foot. 
Coarse aggregate, 110 pounds per cubic foot. 
Cement. 94 pounds per cubic foot. 
Percentage of fine aggregate in new mixture=60. 
1. 1:2:3^=94:2 times 95: 33 times 110. 
=94:190:385 proportion by weight. 
=1:2.02: 4.10=1: 6.12. 
2. Percentage of sand in base mixture aggregate= 
190 
190+385 tinM ' s 100=83 
Percentage of sand in new mixture=60. 
3. 60—33=27. 
2? times 0.05=1.35 
0.12 — 1.35=4.77, therefore the new proportion by weight is 
1 : 4.77. 
4. Since 60 per cent of the total aggregate is to be fine aggre- 
gate. 
1:4.77=1:2.80: 1.91 
5. 1:2.86:1.91=94 pounds: 269 pounds: 180 pounds. 
Dividing by the unit weights of materials. 
94 269 180 
94 " "95" * 110 = * CUD * C f,,ul : 2-83 cubic feet : 1.64 cubic feet. 
Or if unscreened gravel is to be used, determine the unit 
weight of the gravel. Assume for this case 115 pounds per 
cubic foot. 
Then 1 : 4.77=04 pounds to 441) pounds. 
Dividing by unit weights, 
94 449 
94 : ri5 = " 1 cubic foot: 3.90 cubic feet 
This method applies to either screened or unscreened mixtures 
of sand and gravel in which the specific gravities of the fine 
and coarse aggregate are approximately equal, and to percentages 
of fine aggregate greater than 33. 
Note. — The method can bo applied to mixtures of materials of 
different specific gravities, such as sand and crushed limestone, by 
reducing the proportions to relations by absolute volume and mak- 
ing the adjustment upon that basis. 
Recommended practice in the use of unscreened gravel. 
1. The fine aggregate portion must pass the user's specifica- 
tion for fine aggregate. 
2. The coarse aggregate portion must pass the user's specifica- 
tion for coarse aggregate. 
3. Unscreened gravel should no1 be \\H'i\ unless in the opinion 
of the engineer in charge, the amount of sand in the aggregate 
will nol vary more than 10 per cent. 
4. Unscreened gravel should only be used from stock piles con- 
taining the largest practicable amount of material. 
Reference. — For a theoretical discussion of the method of pro- 
portioning see Bulletin No* »'•<>. Engineering Experiment station. 
b»wa State College. For verification sec paper by R. \Y. Cnim, 
Proceedings American Society for Testing Materials. R)22. 
COMPRESSION TEST OF CONCRETE. 
(A. S. T. M. tentative method, serial designation C 39-21 T. modified.) 
1. These methods are intended to cover compression tests of 
concrete made in a laboratory where accurate control of quan- 
tities of materials and test conditions is possible. They are de- 
signed to apply primarily to hand-mixed concrete, These methods 
