The Orange River Ground Moraine. 117 
tunately they only extend some 8 feet from the surface. At both 
these localities the boulders are enclosed in a shaly matrix, which 
is distinctly bedded, as if they had been dropped into their positions 
by ice floating in water covering a muddy bottom. The northern- 
most of the two localities is about 30 miles south of the till west of 
Prieska. 
It is impossible at present to state the maximum thickness of the 
conglomerate, but we think it must be some hundreds of feet. If 
certain patches of shale at Groot Modder Fontein were never covered 
by another band of conglomerate, at that particular locality the 
thickness is small, probably under 30 feet. In view of the fact that 
shales containing few boulders were found at Roode Poort and 
Holgat’s Fontein, some 10 miles from and at a higher level than the 
nearest outcrop of old rocks on Omdraai Vlei and T’Kuip, it is unsafe 
to conclude that the isolated patches of shales, which show no 
boulders in the limited portion seen, are really lying at the top of 
the conglomerate. In other words, the shale may have originally 
lain between two bands of conglomerate. 
We have now to consider the evidence of glaciation offered by 
the rock surfaces underlying the conglomerate. 
At Jackal’s Water, in Prieska division, an outlier of conglomerate 
is underlain by granite and quartzite. On the west is a long range 
of low hills made up of quartzite and quartz schist. The quartzite 
is generally a very hard, massive rock, and at places where this rises 
from under the conglomerate its surface is smooth, rounded, and 
covered with scratches (Plate XIII.). The individual striz are some- 
times 2 feet in length, and cross each other at low angles, but their 
general trend is N.N.H.-S.S.W. (Plate XIV.). From the fact that 
the southern sides of the mounds, which are strictly comparable 
with the ‘‘roches moutonnées’”’ of Huropean geologists, are rough 
and unscratched, while the northern slopes are more gently inclined 
and are smoothed and striated, it is evident that the direction 
in which the ice travelled was from N.N.H.-S.S.W. One such 
‘‘roche moutonnée ’’ rises about 10 feet from the veld, and its long 
northern slope is about 60 feet in length. Plate XIII. shows a group 
of ‘‘roches moutonnées”’ with the boulder-covered ground at its 
foot. These boulders indicate the presence of the conglomerate, but 
the rock itself is not seen in section, although a few small outcrops 
of it are to be met with a few hundred yards off in the same outlier. 
Isolated boulders of rocks foreign to the locality are found scattered 
over the quartzite hills, and prove that the whole area was once 
covered with the conglomerate. 
