6 BULLETIN 105, U. S. DEPARTMENT OF AGRICULTURE. 
Methods and materials. — The same procedure was followed as in the 
construction of experiment No. 4, excepting for the omission of the 
oil from the mixture. From station 33 + 60 to station 42 + 50 this 
section was completed in 1912, but the remainder of the construction 
was carried out in the spring of 1913. 
Hydrated lime experiment. — Between station 49 + 00 and station 
50 + 50, and joining the same experiment in the oil-cement concrete, 
hydrated lime was added to the concrete. The proportions were the 
same as given in experiment No. 4, except the omission of the oil. 
Experiment No. 6. — Vitrified brick. 
Location: Station 52+00 to station 61+SO. 
Area: 2,055 square yards. 
Concrete oase and shoulder. — The concrete base was similar in 
character and construction to that used in experiments Nos. 1 and 2, 
and described in Circular No. 99. The concrete was made up of 1 part 
by volume of Portland cement, 3 parts of sand, and 7 parts of gravel. 
A bag of cement was considered as having a volume of 1 cubic foot. 
The concrete was mixed in a mechanical concrete mixer of the self- 
propelled, rotary-distributor type, the flow from which was practically 
continuous, and the distributor was swung across the road and the 
mixer moved ahead in such manner as to place the concrete upon the 
subgrade in approximately the position desired. It was then shov- 
eled into place in such an amount as to form when compacted with 
concrete tampers a uniform layer 6 inches in thickness. 
Before the concrete base had taken initial set,, a concrete shoulder 
8 inches wide and 6 inches high, mixed in the proportions of 1 part of 
cement, If parts of sand, and 3 parts of gravel was placed upon all the 
edges of the base except those which abut against the concrete gutter. 
Sand cushion. — The cushion sand was spread over the base to an 
approximate depth of 2 inches and struck off to a true depth of 2 
inches by the use of a "strike board," the ends of which rested 
upon the concrete curbs. The cushion was then roiled with a lawn 
roller weighing 300 pounds. After the first rolling, sand was spread 
over the rolled surface to a depth of about one-fourth inch, and 
again struck off and thoroughly rolled. A mechanical analysis of 
the sand used is given in Table 6. 
Table 6. — Mechanical analysis of the cushion sand. 
Per cent. 
Passing ^-inch, retained on 10-mesh screen 7. 1 
Passing 10-mesh, retained on 20-inesh screen 13. 7 
Passing 20-mesh, retained on 30-mesh screen 18. 9 
Passing 30-mesh, retained on 40-mesh screen 20. 5 
Passing 40-mesh, retained on 50-mesh screen 16. 8 
Passing 50-mesh, retained on 80-mesh screen 13. 7 
Passing 80-mesh, retained on 100-mesh screen 1.4 
Passing 100-mesh screen 11 
