Pratt] CORUNDUM GEMS. 41 
box (b of PL II, B), where the large pieces of rock and bowlders arc 
removed and most of the dirt and fine gravel is washed out. They 
are then shoveled into a rocker (c of PI. II, />), where they are further 
cleaned and concentrated, the final concentration of the rubies being 
done by hand. 
No basic magnesian rocks or serpentine derived from them, in which 
most of the corundum of North Carolina occurs, have been found in 
this valley. Corundum Hill and the Klli.jay corundum region are, 
however, less than LO miles to the south. 
Although, in many respects, the occurrence of the rubies and their 
associate minerals in the Cowee Valley is similar to the occurrence of 
he ruby in Burma, no limestone has been found near the alluvial 
deposits, the nearest point a1 which limestone has been found being 
at Cullowee Gap, about 8 miles to the southeast. 
The country rock of the district is a gneiss, of a gray, fine-grained 
variety, which has a great many small garnets disseminated through 
it. The rock for the most part is in a highly decomposed condition, 
but there arc small exposures of the undecomposed rock in many 
places. The gravels in which the rubies arc found rest on a soft rock 
known as saprolite, which is the result of weathering of the basic 
silicate rocks in place. By means of shafts ami the workings at I he 
gravel washings it has been shown that al a depth of .'!■"> feel or more 
these saprolitic rocks contain fragments of the undecomposed rock 
and pass into such rocks as eclogite-amphibolite and a hornblendic 
gneiss. 
A narrow dike of hornblendic eclogite a few feet in width is 
exposed near the present workings of the company and can he traced 
for about K><> yards. 
No rubies have been found in the undecomposed rock, bin at In 
Situ Hill small rubies of a rather pale color were found in a narrow 
band of saprolitic rock. This band was. however, cut off by slick- 
enslides so that it could not be followed in any direct ion. There 
are four parallel slickenslides that have been exposed at one place in 
the workings, t he general direction of the slides being N. 75° E. Some 
of these are 70 feet in length and of unknown depth. It is very evi- 
dent that there has been a great deal of disturbance in this immediate 
vicinity through the breaking of the rock masses by faulting, the 
ready influx of water having caused the reduction of the rocks to 
their saprolitic condition. 
In washing the gravels and masses of saprolite, masses of unde- 
composed rock have been uncovered, and in the center of these 
nodules of the pure hornblende rock have been found. The saprolite 
bordering these nodules often contains particles and crystals of 
corundum. 
Less than 2 miles to the east of In Situ Hill, beyond Betts Cap of 
the Cowee Mountains, corundum of a gray to bluish color, but highly 
