34 BULLETIN 586, U. S. DEPARTMENT OF AGRICULTURE. 
Experiment No. 17 — Bituminous macadam, granite with fluxed native 
asphalt; experiment No. 18, Bituminous macadam, granite with oil- 
asphalt X. — In the spring of 1916 small areas, distributed quite gen- 
erally over the surface of both experiments, began to disintegrate so 
that numerous penetration patches were necessary. These sections 
had been laid over old macadam which was scarified and dressed 
with a layer of new stone before the penetration course was applied. 
The breaking of the newly constructed surface appeared to be due to 
movement in the wearing surface caused by the failure to fill and 
bond the new stone added to the foundation. Both sections were in 
good condition for retreatment and this was applied. The charac- 
teristics of the bituminous material used in the reseal are given in 
Table 37 and the cost data and quantities will be found in Table 38. 
Table 37. — Analyses of residual asphaltic petroleum. 1 
General characteristics, viscous, sticky fluid. 
{A n 991 
B-a989 
Flash point ° C 100 
Burning point ° C 120 
Float test 50° C. (seconds) j^~ ^ 
(Jd — d0 
Viscosity Engler 50 c. c. at 100° C, specific 14. 2 
Loss at 163° C. , 5 hours, 20 grams, per cent 14. 30 
Character of residue very viscous, sticky. 
Float test on residue 50° C. (seconds) 160 
Per cent of total bitumen insoluble in 86° B. , naphtha 17. 39 
Per cent of fixed carbon 10. 44 
Bitumen soluble in CS 2 (total bitumen), per cent 99. 89 
Organic matter insoluble, per cent 0. 11 
Inorganic matter insoluble, per cent - 0. 00 
100. 00 
Experiment No. 19 — Bituminous concrete, granite with oil-asphalt 
X. — This experiment showed marked deterioration and many pene- 
tration patches had to be made in the spring of 1916. It was 
brought into good condition for treatment and a new seal coat 
applied in May and June, 1916. The same bituminous material was 
used as on experiments Nos. 17 and 18, and cost data and quantities 
are given in Table 38. At the time of inspection this experiment 
showed a few wavy spots due to fat penetration patches under the 
retreatment ; otherwise the experiment was in good condition. 
Experiment No. 20 — Bituminous concrete, granite with fluxed native 
asphalt A. — Except at the south end, this experiment did not disin- 
i Samples A and B from each of 2 barrels, except where indicated tests were made on composite 
sample. 
