218 FANNING AND EDDINGFIELD. 
filled with black sand concentrates, forming a compact, smooth 
surface over which the sands flow, carrying away an enormous 
percentage of concentrates and causing the loss of a large amount 
of fine gold. Mercury is not used in the riffles. Some of the 
tables near the screen are cleaned up once a day; others are 
cleaned once a week, and the tables farthest from the screen 
are cleaned only about once a month. These concentrates are 
collected in a box and are screened through a punched metal plate 
with round openings about 6 millimeters in diameter. The 
material passing through the screen flows over a well of mercury 
to catch the amalgamable gold, and from there to a few riffles 
where unamalgamated gold, pieces of amalgam, and mercury 
are caught. This operation recovers the major portion of the 
free gold, a large percentage of the coarse gold bound up with 
quartz, only a very small percentage of the very fine gold, and 
practically none of the fine gold bound up in quartz and pyrites. 
The cleanup concentrates, after receiving this treatment, gen- 
erally are saved for future treatment. 
SAMPLING AND ASSAYING. 
Lot No. 1.—This represents the black sand concentrates from 
which the gold had been collected in the regular cleanup by 
hand-panning without the use of mercury. The gold was sup- 
posed to have been completely panned out, but its abundance as 
revealed in our tests showed that hand-panning is very inef- 
ficient. 
The lot, which weighed about 100 kilograms, was cut into 4 
sections of 25 kilograms each. Section I was cut into halves 
and one of the halves was pulverized through 60-mesh and then 
cut into 4 parts. Each part was divided into 2 splits crushed 
on a bucking board through 150-mesh. The splits representing 
the 4 parts gave the following average assays per ton: 235.24 
pesos,” 210.26 pesos, 217.26 pesos, and 203.80 pesos. Section II 
was similarly treated, except that the half pulverized through 
60-mesh was cut into 2 splits which were independently cut 
down and put through 150-mesh. The samples representing the 
2 splits gave average assays of 299.48 pesos and 306.74 pesos, 
respectively. Comparing the two sections, we find an average 
of 216.88 pesos for section I and an average of 303.10 pesos for 
section II. 
As these checks were very unsatisfactory, a sample of the 
original was bucked through 150-mesh; the assays of this varied 
‘One peso Philippine currency equals 50 cents United States currency. 
