-2 BULLETIN 257, U. S. DEPARTMENT OF AGRICULTURE: 
the 1913 experiments at Miami and two water-gas tar products - 
donated by the manufacturers for this purpose. The oils had been 
stored without protection from the weather, and were in pocr con- 
dition, principally through the presence of water. 
The excellent results obtamed in experiment No. 6 of the original 
series of experiments at Miami could be attributed beyond question 
to the use of screened rock instead of the ‘‘run of the bank” material 
used in the other experiments. Consequently only the screened-rock 
type of construction was followed in these experiments. The rock 
was quarried from the Lemon City pit, which was the nearest source 
of supply. It was comparatively soft and decidedly unsuitable for 
bituminous work of this nature, due to its tendency to crush to pow- 
der when the screened-rock course was even lightly rolled to make it 
conform to cross section. In some portions this created a surface © 
condition similar to that existing in experiments 3, 4, and 5 at Miami, 
described in the previous report. This soft quality and the presence 
of a considerable proportion of fine material, which it was found 
impracticable to screen out, made the surface very susceptible to the 
binding effects of rain. 
The work began at station 0+00 at the north end of the project 
on January 6 and continued intermittently until February 7. 
The general method consisted m scarifymg and regrading the old 
surface, after which a course of new screened rock was spread to a 
loose depth of 24 inches. This course was shaped by a very light 
rolling before spreading the bituminous material over it. The excess 
of bitumen was absorbed by a covering of sufficient fine white sand 
from near-by pits, and the road was finished by thorough rolling. 
Frequent rains and showers durmg the work considerably delayed 
its progress and produced surface conditions which prevented the 
best results bemg obtained. It is clearly demonstrated that the 
greatest difficulty to be overcome in any future large-scale construc- 
tion of bituminous coralline-rock roads is the harmful effect of rains 
on the prepared surface before the application of the bitumen. The 
high cementing power of the rock causes it to bind strongly when 
wet, unless the dust is entirely removed in the screening. The sur- 
face is thus rendered more or less impervious, and the treatment 
becomes in effect only a surface treatment. This difficulty may be 
met by performing such work in the dry season of the year, or by 
keeping at all times the prepared surface work and the bituminous 
application well together. 
The equipment was the same as that used at Miami, consisting of 
a 10-ton gasoline roller, scarifier, grader, two 1-barrel heating ket- 
tles, and hand-pouring pots. Analyses of the bitummous materials 
used are given in the tables following. 
