32 BULLETIN 586, U. S. DEPARTMENT OF AGRICULTURE. 
TaBLE 36.—Analysis of refined water-gas tar—Continued. 
Per cent Per cent 
Distillation: Character. by volume. by weight. 
Wroaiter FI L24Ne ER CEM EYEE ED EN SU PEER E 0.0 0.0 
Hirstilight oils (LO? Ch). Lee Sans oe a PA eel 0.0 0.0 
Second light oils (—110°-170° C.).......... Clear. Trace. Trace. 
Heawiy, oils (170°-2709 C.)g. ee) ese ean 2/25 solid 28.0 24.5 
Heavy, ols (270°—3000C") 02 eo ace cee Clear. 10.3 9.0 
1 Bel ln vale rege til yp hae Aaron Sticky; semisolid. 61.7 66.5 
100.0 100.0 
This treatment prevented further deterioration, although experi- 
ment No. 4 has a somewhat distorted cross section at points where 
the heavy sidehill fill has settled. At the time of inspection both of 
these experiments presented a practically uniform appearance. The 
surface was intact, although the seal coat was blotchy, some areas 
were black and leathery where fat areas had been caused by the pene- 
tration patches in the original surface, and other places were brown 
and mosaic in texture. 
Expervment No. 5—Bituminous macadam, gneiss with fluxed native 
asphalt A; experiment No. 6—Bituminous macadam, gneiss with oil- 
asphalt X.—These experiments were somewhat open and porous in 
the spring of 1916, but at the time of inspection were in good condi- 
tion, having sealed themselves satisfactorily. To prevent raveling, 
a few patches had to be made early in the year before warm weather 
caused bitumen to bleed to the surface. The seal coat is now fairly 
uniform, although several mosaic areas were noted, and the surface is 
somewhat rough, owing to irregularities from construction rather 
than to wear. 
Experiment No. 7—Bituminous concrete, gneiss with oil-asphalt X ; 
expervment No. S—Bituminous concrete, gneiss with fluxed native 
asphalt A.—These experiments were somewhat smoother than ex- 
periments Nos. 5 and 6. Not having been so open in the spring of 
1916, the bleeding above referred to produced a more uniform mat, 
and the surface was in good condition, except for slight irregularities. 
Expervment No. 9—Bituminous gravel concrete, pit-run gravel with 
oul-asphalt Y.—This experiment was somewhat rough and uneven, 
owing to irregularities apparently due to lack of uniformity in the 
seal coat; otherwise the experiment was in good condition. Only 
one or two small areas raveled in the spring of 1916, but were repaired 
easily. 
Experiment No. 10—Bituminous gravel concrete, pit-run gravel with 
fluzed native asphalt B.—This was the poorest of the gravel sections. A 
considerable total area, almost all being on the Four Mile Run fill, 
required patching early in 1916, and practically the entire surface 
received a seal coat in intermittent applications. The mixing plant 
used in construction stood at approximate station 95 and much of 
