104 de. j. w. evans on tee eocks of the [Feb. 1906, 



measuring as much as a centimetre across, and no doubt larger 

 specimens occur. It shows in hand-specimens (M 6 a ; specific gravity 

 = 2*61) a considerable amount of quartz occasionally idiomorphic, 

 pink felspar, and mica. Under the microscope porphyritic crystals 

 of microcline-perthite are visible, which sometimes reach a diameter 

 of 4 millimetres. There are also large crystals of quartz and albite, 

 the latter filled with numerous minute enclosures, which show much 

 greater relative retardation than the felspar, and are probably 

 paragonite. The groundmass is an aggregate of more or less 

 rounded crystals of quartz, microcline, and a little basic oligoclase 

 measuring up to half a millimetre in diameter. 



Another specimen (M 6 b), from the same locality, appears mega- 

 scopically to be similar, but with a finer grain. Under the micro- 

 scope it shows abundant quartz, amounting to about half the bulk 

 of the rock, and occurring in numerous small crystals up to about 



3 millimetres, which, more often than not, show indications of 

 crystalline outline. They are frequently included in the felspar, 

 which is partly microcline and partly albite. There are also a few 

 flakes of decomposed biotite. 



Those portions of the rock that were frequently covered by the 

 river were, like many rocks in similar situations in other tropical 

 rivers, coated with a black film consisting mainly of manganese- 

 oxide and ferric oxide. 1 Both the bare gneiss and the blackened 

 portion had a brilliant polish, the result, I imagine, of the action 

 of the wind-borne sand from the beaches and banks which are left 

 dry when the river falls. It must be remembered, however, that 

 the river-water is heavily charged with sand, and the polishing may 



loses this character : instead of being smooth and polished, it acquires a highly- 

 irregular surface. Above it rise dykes of <?.iorite, crags from 3 to 6 metres in 

 heigbt ; while close at hand sink abysses, or the rock exhibits broad and deep 

 erosions, which become navigable channels of the river when the water is 

 sufficiently high. The platform terminates at the river in one of these cliffs, 



4 metres in height, which throughout its extent forms the margin of the river. 

 In the portion left dry by the river, one sees circular cauldrons a metre and 

 more in diameter and depth, and small elliptical excavations of the same size 

 as those observed in the other cataracts. Some of the rock-platforms are of a 

 shining red colour, perhaps due to ferric oxide ; others are of a brilliant black, 

 due to the oxide of the same metal or peroxide of manganese. Here and there 

 again are blocks split longitudinally, which preserve a remarkable parallelism 

 between the faces of the crack, the salient portions of one side corresponding 

 to the re-entrant portions of the other. Some 50 metres below the cataract, and 

 on the same bank, is another wall (pareduo) like that of Araras, formed of 

 superposed rocks of sandstone and gneiss affecting the form of trap-rocks with 

 such closeness that it resembles an old wall in ruins. The texture of the gneiss 

 resembles that of basalt, but the fracture is more conchoidal. -It was this mass 

 that gave the cataract its name. Thence on to the cataract of the Trez 

 Irmaos ; which is 44 kilometres distant, the river continues studded with rocks, 

 principally on the left-hand side.' 



I saw no evidence of the occurrence of volcanic rocks at Paredao, or any- 

 thing in the nature of an abyss. Col. Church (' The Route to Bolivia via the 

 Eiver Amazon' 1877, p. 188) says that ' the granitic formation which bars the 

 river is very coarse and highly feldspathic. Large crystals of feldspar, several 

 inches in length, are disseminated through the rock, and where the floods have 

 worn them they are as smooth as glass.' 



1 See post, p. 121. 



