The article concerns the results of investigations of the surface relief of quartz grains in two samples of deposits coming from the upper part of the Amazon catchment area, presented in conditions of the natural environment prevailing nowadays and in those of the Pleistocene epoch. The analysis of the greatly magnified sand grains surface was done in a scanning electron microscope of the ISM-25 Jeol type, in the Institute of Geology, Warsaw. The samples of deposits were collected during field investigations in Peruvian Amazonia, in the summer of 1980. The first of the samples comes from the Huallaga valley (right tributary of the Maranón) near the town of Tingo María. It is an area of a warm and humid climate (over 3000 mm of precipitation per year) with intensive processes of chemical weathering, especially among limestone mountain massifs strewn with caves. Even in close proximity of Tingo María the climate is much differentiated; in the neighbouring Huánuco valley, crossed by the Huallaga river, precipitation does not exceed 400 mm yearly and the mountain-slopes are not densely covered with vegetation. The bottom of the Huallaga valley, the mid-channel and coastal sand-banks are built of thick, ill-sorted material containing pebbles, gravel and various-grain sands. The same material may be observed in the exposures of the river’s gorge section (between Huánuco and Tingo María), where it forms high terrace horizons now intensively destroyed by stone-falls and landslides. A cold and humid climate prevailed in the Andes in the glacial epoch. The Central Andes were then glaciated twice at least, and the traces of glaciation such as moraines and post-glacial lakes are common in the upper sections of river valleys of that region. Glaciers developed probably also in the Huallaga valley (above Huánuco) and in the upper parts of the valleys of some of its tributaries. It was probably at that time that the valleys of the Central Andes were highly filled up and among them the Huallaga valley in its gorge section. The analysis of the surface relief of quartz grains in the scanning microscope, by 400-1500 x magnification, has allowed to recognize the character of grain-forming processes and the order of their action. Smooth surfaces have conchoidal breakage with bunches of steps and block relief (Pl. I, photo 4, Pl. II. photo 1). On weathered surfaces there occur fragments covered with a crust (Pl. II, photo 2) which is an indication of dry climate in which silica and other minerals could be precipitated on the grain surface. Other fragments show forms directed in accordance with the crystallographic lattice which points to chemical etching processes (Pl. II, photo 3). There occurs sporadic „lichen” - type chipping (Pl. II, photo 4) observed in areas of intensive mechanical weathering. If more magnified (1500-5000 X) the surfaces modelled by weathering processes display deep corrosive pits (Pl. III, photo 1) frequently having a crystal-directed form (Pl. III. photo 2) and secondary crystallization on the surface. The alternation of the relief types allows to infer on the order in which the processes acted. Three stages may be distinguished: 1) intensive weathering in changing climatic conditions (humid, dry; of high and low temperature); 2) mechanical destruction of the grains; 3) renewed processes of chemical weathering. The time interpretation of those facts presents many difficulties. It cannot be excluded that the glacial period is indirectly recorded in the stage of intensive disintegration of quartz grains, separating two periods of active weathering processes. Thus the glaciation period would be preceded by a period of very dry climate and by another - of great humidity and presumably of high temperature. Uncluckily it is impossible to infer on the succession of those processes by the remaining forms of relief. The post-glacial period was marked only by chemical weathering. However, another origin of the fresh structures preserved on grain surfaces cannot be excluded. In strongly dynamic rivers, carrying masses of pebbles, quartz grains may undergo grinding and trituration like in rock mills. Thus the sharp-edged forms preserved on grains may be as well as relict of the time of flow of fluvioglacial water as they may be shaped at present in highly dynamic rivers to which the Huallaga belongs. The other sample of deposits comes from the Amazon Lowland. It was collected from the edge of a mid-channel sand-bank in the Amazon valley, near Iquitos. The place of collection lies ca 130 km below the confluence of the Ucayali and the Maranón. In this part of the Amazonia a humid, equatorial climate prevails, the mean temperature being 25°C and the precipitation - 2500 mm yearly. It is a low-lying and relatively flat area. The small differentiation of relief is connected with a relatively simple geological structure. Vast plateaux are built of terrigenic formations of the Upper Tertiary (Pebas Formation) and of younger deposits dating from the end of the Tertiary and the beginning of the Quaternary (Iquitos Formation) filling the valley-like depressions cut in older formations. Quaternary and contemporary deposits fill the valleys and build terrace horizons. It is very difficult to define precisely the age of Quaternary acumulation horizons. The phase of intensive erosion in the Amazon valley which descended, along the section from the river mouth to Manaus (1500 km), much below the present-day sea level, is connected with the period of pre-Flandrian regression corresponding with the Vistula glaciation. The climate of the present-day Amazonia was then drier than nowadays. In place of the now existing forests (selva) there grew a savannah with small spots of forests in more humid habitats. At that time, in the Andes foreland, there may have occurred the accumulation of large cones and terrace horizons. The formation of the flood-plain terrace (3-8 m high) in the Ucayali and the Amazon valleys can be probably connected with that period. Similar conditions must have prevailed in the farmer glaciation (the Riss) and the above-floodplain terrace (10-12 m high) could have been formed then. The rise of the ocean level in the Holocene (Flandrian transgression) and a more humid climate caused the expansion of forests and the set-back of erosion. A negative balance of alluvia due to dense vegetation did not allow the deposits to fill the Amazon valley in its lower part; moderate erosion cannot be excluded in the river’s upper part. Among the present-day processes modelling the relief in the river valleys, beside fluvial processes, the activity of the wind should be mentioned, which makes the water wave. This process is favoured by the considerable width of rivers and the existence of numerous overflow arms. The results of waving, such as beaches and cliffs reaching 1.5 m of height, can be seen on both sides of the river bed. Strong waving is also caused by the movement of boats. The analysis of quartz grains of the second sample has shown essential differences in the character of the relief forms of the grain surface as compared with the sample from the Huallaga valley. By 400 X magnification monadnock relief connected with weathering processes, and smooth, rounded edges (Pl. IV, photo 3) may be observed. Larger magnification (2000-7000 X) allows to interpret the relief type in its genetic sense and to connect it with appropriate processes. The oldest process recorded on the grain surface was that of precipitation of amorphous silica forming crusts (Pl. IV, photo 4, Pl. V, photo 1). It may be the record of weathering processes in dry, continental climate of the last glaciation. Another type of forms occurring on all the grains under investigation are V-shaped cuts chaotically disposed (Pl. VI, photos 3, 4, Pl. VII, photos 1, 2), characteristic of beach environment. Some of the V-shaped forms have developed on crust-covered surfaces, some - on fresh surfaces. At times accumulations of V-shaped forms may be observed together with crescent-shaped cuts (Pl. VII, photos 3, 4) which might indicate beach environment where intensive processes of chemical weathering cooperate. It may be supposed that both types of forms (forms connected with chemical weathering and forms developed through abrasion caused by waving) develop nowadays in the region of the upper Amazon. Waving and great changes of water level over the year cause strong side-erosion of higher terraces. This may explain the occurrence, in the alluvia, of crust-covered grains which probably lie on a secondary deposit.