DUST PREVENTION AND ROAD PRESERVATION. 27 
pie ted when a heavy rain fell and on the following morning the sand 
mat was found washed into ridges and to a large extent carried off 
the road entirely. The portion of the experiment on which the oil 
had stood the shorter time became rather mushy and uneven, while 
the remainder of the section was fairly smooth and hard. The section 
was resanded but the same condition was evident two weeks later. 
Experiment No. 3. — Penetration method — Hot, heavy oil. 
Location: From station 5+49 to station 7+49. 
Length: 200 feet. 
Method. — In this experiment, after the old surface had been 
regraded. a course of new unscreened rock was spread to a loose 
depth of 2J inches. This course was brought to the required shape 
with a very light rolling in an effort to permit penetration by leaving 
the crust as open as possible. Light rains previous to the application 
of oil caused the surface to become slightly water-bound. The hot 
oil was applied at the rate of approximately 0.6 gallon per square 
yard, but owing to the varying texture of the surface it readily pene- 
trated on some areas and remained almost entirely on the surface of 
others. The surface was sanded and rolled as in the previous 
experiments. 
Experiment No. 4. — Penetration method — Cold, light oil. 
Location: From station 7+49 to station 9+16. 
Length: 167 feet. 
Method. — This experiment was quite similar to No. 3, in that the 
application was made on a 2^-inch course of new, unscreened rock. 
The rock was, however, heavily rolled in an effort to tighten the sur- 
face sufficiently for brooming the oil. This was partially effected and 
the penetration seemed to be as go6d as in the more lightly rolled 
section. Oil was applied at the rate of 0.7 gallon per square yard and 
allowed to stand for three hours before sanding, and at the end of 
this time it was practically all absorbed. 
Experiment No. 5. — Penetration method — Hot, heavy oil. 
Location: From station 9+16 to station 11+55. 
Length: 239 feet. 
Method. — This experiment was essentially the same as experiment 
No. 4 except for the reason that the course of new, unscreened rock 
was reduced to a loose thickness of 1^ inches in an effort to reduce the 
cost of this type of experiment. From station 9 + 16 to station 10 + 
87 sand was used for a top dressing. The surface cracked and lifted 
under rolling, and required patching in several spots a few days later. 
From station 10 + 87 to station 11+55 new rock screenings were used 
for the top dressing. 
