A Móra Ferenc Múzeum Évkönyve, 1970. 1. (Szeged, 1970)

for the admixture of sands blown out from the flood-plain during the last loess deposition. Consequently, they prove that the river must have appeared already earlier here. The loesses on the western side of the lake are not sandy because this bank of the river was steeper and the winds could blow out sands only from the flat floodplain of the eastern bank (Fig. 8, C/l). Below the loess it is medium sand that predominates throughout the area (Fig. 5, 8, C). In Holocene time the river abandoned its channel, where an ox-bow lake was formed. In the water-covered part of the lake the fine-grained, chiefly clayey, sediments of this ox-bow lake can already be found. As for its origin, Lake Fehér of Kardoskut was formed in the same way, though possibly a little later (B. Molnár and M. Mucsi 1966). The identity in origin is well demonstrated by a comparison of single sections of the two lakes. As evident from Fig. 6, the two profiles are totally the same in character, the only difference consisting in that Lake Fehér is the result of the combination of two ox-bows. In the eastern, narrower, part of the lake used to be a larger channel, in the western part a smaller one (Fig. 4). This is proved by differences in the lithology of the channelfilling material, for the eastern channel contains essentially coarser sediment, mainly medium-grained sands (B. Molnár and M. Mucsi 1966). The present-day, greater, width of the western part is due to underwash and resultant collapse of the river bank there. NATRON LAKES OF THE TISZA VALLEY Of these, Lake Dongér at Pusztaszer has been investigated in detail (B. Molnár, M. Mucsi and L. Magyar 1969) (Fig. 3, 7). An important key to the understanding of the origin of the lakes here may be provided by the loess layer present. According to the results of M. M.-Faragó, it seems to have been deposited under a warm climate, probably in the W

2

—W

3 interstadial. During the last stadial, W

3

, the amount of precipitations decreased, so that fluvial accumulation was interrupted, being replaced by deposition of the aforementioned loess layer. Both the formations, the fluviatile-flood-deposited sequence below the loess and the loess itself, transgress over the present-day eastern boundary of the Danube-Tisza Interfluve and continue westwards under the Holocene windblown sands, too. Early in Holocene time, under a climate becoming warmer and wetter, rainfall grew more abundant, the flood-plain of the Tisza widened again, and the river dissected the loess deposits of W

3

, wearing away their upper part in many places. The sediments deposited that time can also be traced for a little destance westwards under and within the wind-blown sands, as shown by Fig. 3, Section II, Borehole 15. In the dry hazel-nut phase of the Holocene, with overall wind-blown sand movement, these sands migrated southwestwards here too, in accordance with the predominant wind direction, enveloping a part of the Lowermost Holocene flooddeposited sediments, too. As a result of the post-hazel-nut, more humid, climate, stagnant waters left over by the frequent floods have been preserved. Farther away from the Tisza, only fine-grained sediments have been deposited in a water of low energy, and the rate of accumulation was slowed down. In the near-channel zones, however, the river had greater accumulating power. Therefore, the areas farther off have with 72