ORES OF GOLD 61 he pyrites are angular and not water-worm, and that the rounded pyrites might be concretionary. Tt was also claimed that as the Rand gold contains from 100 to 120 parts per 1000 of silver, it cannot be alluvial. "On these grounds the gold was attributed to Ventersdorp age, when the Rand beds were intersected by dykes, against which a patch of rich ore may end abruptly. (The latest full statement of this Sho is by C. B. Horwood, Gold Deposits of Rand, 917. These arguments are however inconclusive. Gold in blacer deposits occurs in the cement and not in the pebbles, which represent the hard barren “ huck-quartz.” Some placer gold, as in Queensland, contains 50 per cent. of silver and is of much lower grade than that of the Rand. Theangu- larity of the gold is due partly to its having been squeezed between the grains of sand, and partly to the gold having been P16. 22,—A PyriTic PEBBLE FROM THE BANKET, A pyritic pebble from the Banket with wind-shaped surfaces, pseudomorphic after 2 pebble of iron oxide. The pebble ic half-an-inch in width. dissolved and redeposited. The sudden ending of rich Patches against a dyke is due to its rise along a fault; the abrupt termination of the patch is due to the fault. The placer theory of the Rand was faced by two difficulties —the rarity of pyrites in ordinary placers, and the rich concentration of such minute particles of gold. Pyrites occurs in placers containing abundant vegetable matter, but 20t in quartz sands where it would be destroyed by oxidation. The Rang pyrites however often occurs in streaks and patches like black iron sand on a sea beach; the Rand pyrites was Probably deposited as iron oxide. Some of the larger pebbles of pyrites (Fig. 22) have sand-worn faces, and must have been originally pebbles of either ironstone or quartz which has been ‘placed by pyrites. X Gold in particles as minute as those in the Rand occurs in sands ang silt, but rarely amounts to more than a penny- weight or so per ton: whereas some Banket contains Over