Dr. John Wm. Evans — Recent Breccias in Bolivia. 551 



Beyond is nothing, except the forest with its smooth level floor 

 of vegetable mould covered with undergrowth, scarcely a hollow 

 to represent the river-bed.^ Indeed, in the lower portion of the 

 course of the stone river the fragments are in some places piled up 

 higher than the adjoining forest soil. 



The other rivers are similar in character." In some cases the 

 width of the bed of rock-fragments is much greater. Occasionally 

 it may divide into branches, and these may again unite so as to 

 surround an island of forest. The larger streams do not disappear, 

 but continue beyond the end of the breccia as ordinary rivers. 



In times of heavy rainfall the water from the mountains comes 

 pouring out through the gorges, hurrying along with it the rocks 

 that have fallen into the river-channels. When the flood reaches 

 the forest slope, the fragments are carried on for a distance that 

 depends on their size and the strength of the current. Eocks at least 

 a foot in diameter may be deposited even near the end of the river 

 of stones, for the forest and undergrowth usually keep the solid 

 material as well as a great part of the water from diverging to one 

 side or the other, although there may be no banks to keep them in. 



When the plain, already flooded by the rains, is reached, the river 

 spreads out, loses its current and identity as if it had entered a sea or 

 lake, and leaves the last of its burden of sandstone blocks behind. 



Much of the transport, especially of the larger fragments, takes 

 place at the time of exceptionally great floods, and it is probable 

 that some blocks may lie in the gorges for years till an abnormal 

 downpour gives the rivers sufficient force to sweep them away. 



After a river-bed has been raised by successive accumulations 

 of stone fragments, the time comes when a strong flood finds it 

 easier to take a new course in spite of the obstacles presented by the 

 forest. Usually the river extends its bed on one side, desti'oying 

 the trees that stand in the way, and leaving part of its former 

 course to be overgrown by vegetation, or it may take an entirely 

 dififerent path and gradually send out a new river of stones into the 

 forest. 



By changing their courses in this way from side to side, the rivers 

 slowly raise the level of the forest slope, with the assistance,^ near 

 the foot of the mountains, of the smaller streams that flow off the 

 side of the outer hills. 



At some points the rivers cut through former accumulations of 

 breccia, so that there is an opportunity of examining its structure. 

 This appears to be identical with that of the material now being 

 laid down. As would be expected, there is not much trace of 

 stratification, though there is a tendency for the fragments to lie 

 with their longer axes horizontal. There are few remains of the 

 forest, for wood enclosed among the fragments and exposed to the 



1 I was told that further to the north-east at a lower level — for even the plain 

 has a slight inchnation — the stream reappears at the surface. 



2 On the road from San Buena Ventura to Tumupasa, which runs through the 

 forest for about forty miles parallel to the mormtaius, I passed nineteen stone rivers, 

 of which two were cLiy, besides numerous smaU forest streams without stones. 



