them on his wheel. The discovery of some of these caldeiroes after expensive labors, some thirty years ago, afforded sufficient return to enrich the families of the two partners who worked them. An eyewitness to the fact relates that in one of the holes under the diorite arch, when the superficial layer of sterile sand was removed, a clear mass of the precious stones was revealed, and the discoverers were able on the spot to fill their pockets with diamonds. In the Ribeirão do Inferno, a single caldeiroe of a few cubic metres' capacity furnished nearly 8,000 carats of diamonds. Such fortunes, however, are extremely rare, and can not be counted on in the regular mining.
It is easy to comprehend how powerful must have been the action of the erosion above described, the duration of which is measured by many millions of years, but is quite outside of our chronology. The greater part of the ravine through which the Jequitinhonha flows, as well as the whole groundwork of the hydrographic system of the region, is doubtless due to phenomena of upheaval, the directions of which oscillate around a north-and-south axis. The tributaries of that river have cut out channels for themselves which are now deep, close canons, with sharply cut, precipitous walls.
The almost horizontal disposition of the sandstone strata, and of the conglomerates, which form the banks of the rivers, and the terraces which indicate the successive levels occupied by the river-bottoms, leave no doubt of the correctness of this assertion. At first, consequently, the bottoms of the streams were almost at the level of the surface; the rivers overflowed their banks after light rains, and their diamond-bearing sands were spread over the table-lands and in the mountain-gorges. As the beds of the rivers were worn down deeper in the rock, overflows became more rare, and the sands were carried to only small distances from the banks. Finally, after a certain period, inundations became impossible, and the sand was deposited in the caldeiroes, the caves, and subterranean channels which the river wore out in the rocks over which it flowed. This epoch may be referred to the period which preceded our own, and which is characterized in Europe by the stone implements of human origin and use, specimens of which have been found in the diamond-bearing land.
Then, either by a rising of the coast, or, as is less probable, by a subsidence of the central plateau of Minas Geraes, the fall of the streams diminished, and, instead of continuing to excavate their beds, they began to fill them with the diamondless, shifting deposits, the formation of which is continued into our day. The sands of the table-lands and the banks of the rivers are much less rich than those of the streams; and in any case the diamonds have to be separated, by methodical washings, from the foreign substances with which they are mixed. Very rarely a simple sifting with the fingers suffices to extract the jewel. In 1824 a region was discovered in the lower part of the table-land of Diamantina where the diamonds were scattered